author manuscript
TRANSCRIPT
Serveur Academique Lausannois SERVAL serval.unil.ch
Author ManuscriptFaculty of Biology and Medicine Publication
This paper has been peer-reviewed but does not include the final publisherproof-corrections or journal pagination.
Published in final edited form as:
Title: Molecular Microbiology: Diagnostic Principles and Practice, 3rd
Edition
DOI: 10.1128/9781555819071
In the absence of a copyright statement, users should assume that standard copyright protection applies, unless the article containsan explicit statement to the contrary. In case of doubt, contact the journal publisher to verify the copyright status of an article.
1
“Molecular and Mass Spectrometry Detection and 1
Identification of Causative Agents of Bloodstream 2
Infections” 3
Onya Opotaa, Katia Jatona, Guy Prod’homa and Gilbert Greubab 4
5
Institute of Microbiologya and Infectious Diseases Serviceb, University of Lausanne and University 6
Hospital Center, Lausanne, Switzerland 7
8
9
Corresponding author: 10
Gilbert Greub, 11
Institute of Microbiology, 12
University of Lausanne and University Hospital Center, 13
Bugnon 44, 1010 Lausanne, 14
Switzerland. 15
Phone +41 (0)21 314 4979 16
Fax +41 (0)21 314 4060 17
E-mail: [email protected] 18
2
Sommaire 19
ABSTRACT .................................................................................................................................................. 4 20
INTRODUCTION ........................................................................................................................................ 5 21
1. Bloodstream infections ..................................................................................................................... 5 22
2. Blood culture-based diagnosis of bloodstream infections ................................................................ 7 23
3. Microbial diagnosis of bloodstream infections directly from blood ................................................. 8 24
MOLECULAR METHODS FOR THE IDENTIFICATION OF PATHOGEN FROM POSITIVE 25
BLOOD CULTURES ................................................................................................................................. 10 26
1. Fluorescent in situ hybridization (FISH) ........................................................................................ 11 27
2. Microarray ....................................................................................................................................... 12 28
3. Nucleic acid amplification-based methods ..................................................................................... 14 29
Multiplex PCR .................................................................................................................................... 14 30
Rapid PCR-based systems .................................................................................................................. 15 31
4. Matrix-Assisted Laser Desorption Ionization Time-Of-Flight Mass Spectrometry (MALDI-TOF 32
MS) 16 33
Principle of MALDI-TOF MS ............................................................................................................ 16 34
Application of MALDI-TOF MS on positive blood culture bottle ..................................................... 17 35
MOLECULAR METHODS FOR THE DETECTION OF PATHOGENS DIRECTLY FROM BLOOD 20 36
1. Broad range PCR ............................................................................................................................ 21 37
2. Real-time PCR and Multiplex PCR ................................................................................................ 23 38
3. PCR-Electrospray Ionization Mass Spectrometry (PCR/ESI-MS) ................................................. 29 39
Principle, characteristics and performance of PCR/ESI-MS .............................................................. 29 40
Application of PCR/ESI-MS to the diagnosis of BSI directly from blood ......................................... 31 41
MOLECULAR DETECTION OF ANTIMICROBIAL RESISTANCE DURING BLOOD STREAM 42
INFECTIONS ............................................................................................................................................. 34 43
1. Rapid PCR Based-methods for resistance detection from positive blood cultures ......................... 34 44
Detection of resistance mechanisms of gram-positive bacteria .......................................................... 34 45
Detection of resistance mechanisms of gram-negative bacteria ......................................................... 36 46
2. Resistance detection from positive blood culture using MALDI-TOF MS .................................... 37 47
3. Resistance detection directly from whole blood ............................................................................. 38 48
CONCLUSIONS AND FUTURE DIRECTIONS ...................................................................................... 38 49
REFERENCES ........................................................................................................................................... 43 50
TABLES AND FIGURES .......................................................................................................................... 80 51
3
52
53
4
ABSTRACT 54
55
Bloodstream infections (BSI) are severe diseases associated with a high morbidity and mortality, which 56
increases with the delay until the administration of the first appropriate antibiotic. For this reason, 57
empirical treatments, made of broad spectrum antibiotics, are rapidly started when a BSI is suspected on 58
the basis of the clinical and epidemiological data, but this does not exclude any risk of inappropriate 59
initial treatment and make the microbial diagnosis crucial. The time to positivity of blood culture, 60
currently the gold standard to establish the etiology of bloodstream infection can vary from few hours to 61
several days and using conventional (culture-based) methods, one or more days can be necessary to 62
determine the etiologic agent of a BSI. Recent advance in molecular biology have permit to develop new 63
methods to accelerate the microbial diagnosis of BSI. Non-amplified nucleic acid-based methods such as 64
fluorescent in situ hybridization (FISH) or microarray can be used on positive blood culture whereas 65
amplified nucleic acid-based methods such as (real-time) PCR or PCR-Electrospray Ionization Mass 66
Spectrometry (PCR/ESI-MS) can be used directly on whole blood. Matrix-Assisted Laser Desorption 67
Ionization Time-Of-Flight Mass Spectrometry (MALDI-TOF MS), based on the analysis of the mass 68
spectrum generated by bacterial proteins can be used on positive blood culture. We will present these 69
technologies and their performance; we will also discuss their advantages and their inconvenient as well 70
as the question that need to be addressed to fully benefit from these new technologies. 71
72
73
74
75
76
5
INTRODUCTION 77
78
1. Bloodstream infections 79
Bloodstream infections (BSI) are severe diseases associated with a high morbidity and mortality, which 80
increases with the delay until the administration of the first appropriate antibiotic (58; 82; 90; 113; 148; 81
149; 157; 232). For this reason, empiric treatments made of broad range anti-infectious compounds or 82
made of a combination of antimicrobials are started immediately after the sampling of blood bottles. BSIs 83
can be caused by various microorganisms. In the absence of microbiological documentation, a BSI is 84
suspected by the physicians on the basis of clinical symptoms, which trigger the start of empirical 85
treatments. The clinical presentations are multiple and include: fever or hypothermia, increases in heart 86
rate, change in inflammatory variable (C-reactive protein, procalcitonin and white blood cell count 87
increase), organ failure (58; 178). These symptoms are generally non specific and only suggest 88
bloodstream dissemination. Empirical treatments are made of broad spectrum antibiotics on the basis of 89
the clinical and epidemiological data, but this does not exclude any risk of inappropriate initial treatment. 90
Similar to delayed introduction of the first antibiotic treatment, an inadequate treatment is associated with 91
a significant increase of mortality (188; 265). In a recent study there was a nearly 3-fold increase in the 92
risk of mortality when the antimicrobial treatment was inappropriate (28). Pseudomonas aeruginosa or 93
Acinetobacter baumannii that are associated with frequent multi-drug resistant profile are often the 94
microorganisms for which empirical antibiotic coverage are inappropriate. Enterobacteriaceae producing 95
Extended Spectrum Beta-Lactamase (ESBL) such as Escherichia coli or producing inducible 96
cephalosporinase such as Enterobacter cloacae, are also commonly associated with an inadequate 97
empirical treatment (37). The mortality rate of patient suffering of a methicillin resistant Staphylococcus 98
aureus (MRSA)-associated BSI was significantly higher among patients receiving an inappropriate 99
empirical antibiotic treatment (168/342, 49.1%) than among patients receiving an appropriate empirical 100
6
antibiotic treatment (56/168, 33.3%) (92; 187). Inadequate treatments can also occur for organisms such 101
as Enterococcus faecalis species that are intrinsically resistant to some commonly administered first-line 102
antibiotics such as Ceftriaxone (265). Broad-spectrum molecules also have a detrimental impact on 103
beneficial bacteria that constitute the protective flora and contribute to the emergence of multi-drug 104
resistant strains. Moreover, some of these antimicrobial molecules can also have a toxic effect for the 105
patient (144; 188; 265). The rapid identification of the causative agent of the BSI thus allows the 106
adjustment of the anti-infectious therapy or the reduction of the spectrum (de-escalation) with significant 107
clinical benefits. 108
BSIs are characterized by a low quantity of circulating microbes. On the basis of plating methods, the 109
bacterial load has been estimated to be around 1 to 10 CFU per ml of blood in adults (105; 132; 133; 243; 110
251). This method is efficient to determine the amount of bacteria that survived the plating protocol; 111
hence the true number of circulating bacteria is likely underestimated. DNA copies present in the 112
circulation should be even higher as it also includes DNA released by dead bacteria or clumping bacteria 113
as well as DNA from bacteria engulfed in phagocytes. On the basis of quantitative PCR, Bacconi et al. 114
have recently estimated that during bacteremia, the blood contains 103 to 104 bacterial cells per ml (13), 115
which is higher than the analytic sensitivity of most of the available molecular methods. 116
Hence, the ideal diagnostic method for BSI should be: 1) able to identify a broad range of pathogen, 2) 117
have an analytic sensitivity lower than 10-100 CFU/ml, 3) rapid, 4) quantitative to give an idea of the 118
severity of the infection, 5) give an antibiotic susceptibility profile, 6) associated with low hands-on time 119
and 7) automated (65). These are all potential characteristics that molecular diagnosis methods may fulfil, 120
at least to some extent. 121
Molecular methods for the diagnosis of BSI refer to nucleic acid-based methods but also to non-nucleic 122
acid-based. PCR based methods that constitute the nucleic acid amplification-based methods can be 123
applied both on positive blood culture and on whole blood. Indeed the PCR step increases the sensitivity 124
7
of the detection which makes it suitable for paucibacterial samples such as blood (Figure 1)(174). 125
Nucleic-acid based methods non amplification based such as FISH (fluorescent n situ hybridization) or 126
microarray have limited sensitivity. Hence, they are restricted to positive blood-culture. For the same 127
reason, MALDI-TOF MS (Matrix-Assisted Laser Desorption Ionization Time-Of-Flight Mass 128
Spectrometry) a non-nucleic acid based method that analyse microbial protein mass-spectrum is suitable 129
for diagnosis from positive blood-culture. 130
In this chapter we will present the molecular methods available for the microbial diagnosis of BSI. We 131
will report and discuss the performance as well as the advantages and the inconvenient of these methods. 132
2. Blood culture-based diagnosis of bloodstream infections 133
Blood culture (BC) is currently the gold standard to establish the etiology of bloodstream infection. Blood 134
bottles contain specific liquid broth for growth of bacteria or fungi present in blood. Distinct media are 135
available to grow different microbes (174). BC-based diagnosis has been improved by the use automated 136
incubators that can detect bacterial or fungal growth into blood bottles thanks to associated 137
physicochemical variations. When the automated system detects the growth, visualisation of the 138
microorganism is possible using standard Gram or fluorescent staining (figure 2). The Gram staining can 139
give a first presumptive etiology of the BSI. However, only the final identification of the pathogen and its 140
antibiotic susceptibility testing can insure the adequacy of the ongoing antibiotic treatment. Thus, when a 141
blood culture is detected as positive, diagnostic laboratories have to identify the microbe rapidly and with 142
the highest sensitivity and specificity. However, the time to positivity (TTP) of the blood-culture bottles 143
can vary from few hours to 24-48 hours for fast growing bacteria, and even more for slow growing 144
bacteria or fungi. In addition, it is generally admitted that more than 50% of BSI occur with negative 145
blood culture (57; 79). When the blood culture become positive, using conventional methods, one or more 146
days can be necessary to identify the pathogen and to determine the exact etiology of the BSI (figure 2). 147
These methods include phenotypic characterisation using biochemical and enzymatic tests. Most of these 148
8
phenotypic characterizations require a subculture in liquid media or a subculture on solid media to obtain 149
isolated colonies. Finally, BSI involving multiple organisms may also further increase the time to 150
identification of most detection methods. 151
Inside positive BC, bacteria concentration reaches 1.106 to 2.108 for gram-negative cocci and 2.107 to 152
1.109 for gram-negative bacilli (42; 229). These concentrations allow the use of amplification-based 153
methods such as PCR and real-time PCR as well as non amplification-based methods such hybridization 154
or MALDI-TOF MS (figure 1). 155
Blood is a challenging sample for the detection and the identification of pathogens since it contains a low 156
number of microorganisms in comparison to human components (DNA, proteins and cells) (175). These 157
compounds can interfere with the detection or identification by leading to false positive or false negative 158
(inhibition). All these components are transferred into the blood culture at time of the blood sampling. 159
MALDI-TOF MS is a non-nucleic acid based molecular methods that has considerably accelerated and 160
simplified the identification of pathogen from positive culture. MALDI-TOF MS is based on the analysis 161
of the mass spectrum generated by bacterial proteins, mainly house-keeping proteins (figure 3). This 162
technique is now applicable on positive blood culture but still requires a sample preparation step to 163
discard blood and other non-bacterial components. 164
3. Microbial diagnosis of bloodstream infections directly from blood 165
The diagnostic of BSI directly from whole blood has been a major concern for medical diagnostic 166
microbiologists. Indeed blood cultures appear to have some intrinsic limitations: 1) approximately fifty 167
percent of BSI are BC negative (57; 79) and 2) in the case of positive BC, the TTP can vary from hours to 168
days 3) BC requires a high quantity of blood that is difficult to obtain from some patients such as 169
pediatric patients 4) an antibiotic treatment is often initiated prior to blood culture. Many technical 170
improvements have been made to increase the performance of detection methods from whole blood but 171
some limitations remained. Nucleic-acid based methods were limited by the need of large volume of 172
9
blood due to the low number of CFU per ml associated with the presence of human DNA in excess 173
mainly due to white blood cells DNA (4). In addition, these methods are also sensitive to contaminant - 174
bacterial or fungal DNA – or to the presence of DNA from dead organisms that could lead to false 175
positive results (21; 162; 237; 238). Quantitative analyses are more powerful to interpret such positive 176
results. 177
To improve the sensitivity of or nucleic acid based methods, the excess of human DNA should be 178
removed. This could be achieved by removing white blood cells before nucleic acid extraction or by 179
selective removal of human DNA after extraction. Nucleic acid based methods, in particular PCR, are 180
sensitive to various inhibition mechanisms. Some of these mechanisms have been characterized but many 181
of them remain unknown (112; 175; 195; 259). Inhibitory compounds can be contained in the sample or 182
they can be the result of the sample preparation process. The blood contains some well known PCR 183
inhibitors such as haemoglobin, bile salts and heme found in erythrocytes (2; 4; 111) as well as in 184
lactoferrin found in leukocytes (4). The inhibition of PCR-based methods by these compounds is due to 185
the release of iron (from heme) that is known to interfere with DNA synthesis (27). Immunoglobulin 186
present in the bloodstream in particular IgG can inhibit PCR by binding single stranded DNA (3). Red 187
blood cells present in blood bottles can also impair non-nucleic acid based methods such as MALDI-TOF 188
MS as they bring a high amount of proteins that would interfere with the quality of the protein mass 189
spectrum. Several protocols are now available for their removal prior MALDI-TOF MS analysis from 190
positive blood bottles (51; 174; 253). 191
At present, each available molecular method is associated with a specific sample preparation. For the 192
detection of a large number of organisms, the limited sensitivity of broad-range PCR targeting house-193
keeping genes have been overcame by the use of multiplex PCR or multiple real-time PCR. A recent 194
technology, PCR/ESI-MS (PCR and Electrospray Ionization Mass Spectrometry) is a method that 195
associates PCR and ESI-MS. The PCR allows the specific detection of low amount of bacterial, fungal or 196
viral DNA from whole blood. This technology is associated with the analysis of PCR amplicon by ESI-197
10
MS that determine the base composition rather than the nucleotide sequence. This has been shown to be 198
sufficient for organism identification at the species level. Methods are now developed to remove the 199
excess of non bacterial DNA and or inhibitors (4; 195; 202; 259). Most of these methods are time 200
consuming, require intensive labour and are specific for one type of sample, one type of analysis or one 201
type of device. 202
MOLECULAR METHODS FOR THE IDENTIFICATION OF PATHOGEN 203
FROM POSITIVE BLOOD CULTURES 204
BC is currently the reference method for the identification of pathogens involved in BSI because it is easy 205
to perform and sensitive due to the large volume of blood that could be analysed when multiple blood 206
bottles are collected (174). When a BC is positive, conventional methods including phenotypic 207
characterisation and rRNA gene sequencing have an extremely high specificity. However these methods 208
are time consuming because they require an additional subculture and are challenged by a broad variety of 209
molecular methods. Fluorescent in situ hybridization (FISH), the first molecular method that has been 210
developed, is based on the binding of specific probes on pathogens DNA. FISH is a rapid, sensitive and 211
specific method but is dependent on the choice of the probes to be tested. Microarrays allowed the 212
detection and identification of pathogens as well as resistance genes. PCR-based methods allow the rapid 213
detection of a single organism (specific real-time PCR) or multiple organisms (broad-range PCR or 214
multiplex PCR). These methods are highly sensitive to contaminating DNA or PCR inhibitors. MALDI-215
TOF MS, one of the latest commercialized technologies is presented as a revolution for clinical 216
microbiology laboratories (18). Indeed, MALDI-TOF MS may identify in less than one our most bacterial 217
and fungal agents generally recovered from BCs. 218
11
1. Fluorescent in situ hybridization (FISH) 219
Fluorescent in situ hybridization (FISH) is based on the specific binding of fluorescent peptide nucleic 220
acid probes (PNA) to the rRNA – 16S rRNA for bacteria and 18S rRNA for fungi and other eukaryotes. 221
Basically, slides are prepared from a positive blood culture, the fluorescent PNA are applied and unbound 222
PNA are removed by a washing step. Bound fluorescent PNA is generally observed using a fluorescent 223
microscope as flow cytometry detection still need further investigation and development (11). FISH 224
requires a high number of living cells which prevent its use directly on whole blood. Multiples probes that 225
recognize microorganisms at the genus level or at the species level are available. All the solutions are 226
based on multiple probes that detect and distinguish two or three organisms. A first kit of the PNA-FISH 227
system (AdvanDx, Wolburn, MA) allowed the detection of S. aureus and coagulase negative 228
staphylococci (36; 86; 93; 104; 173). A second kit allows the detection of Enterococcus faecalis versus 229
Enterococcus faecium and other Enterococcus spp. (OE) (87). E. coli and P. aeruginosa can be detected 230
using a PNA mix or using a mix of three probes that include also K. pneumoniae (183; 189; 220). C. 231
albicans, C. glabrata, C. parapsilosis and C. krusei can also be detected (74; 100; 199; 215; 255). 232
The Quick-FISH system is a more rapid version of the PNA-FISH system. The turnaround time (TAT) is 233
shortened by reducing the hybridization phases to 15 min and by removing the wash step. At present, the 234
QuickFISH system provides kits that detect and distinguish S. aureus/Coagulase negative staphylococci, 235
Enterococcus faecalis/Enterococcus faecium, E. coli/P. aeruginosa (55; 56; 158). The diversity of 236
microorganisms that is currently detected is still limited but these bacterial species represent about 95% of 237
the pathogen involved in BSI. The choice of the probe to be tested will depend on the Gram result, the 238
clinical presentation and the local epidemiology. The sensitivity of the QuickFISH system is 99.5% for 239
the detection of S. aureus and 98.8% for coagulase negative staphylococci, with a combined specificity of 240
89.5% (55). Martinez et al. reported 97.9% of concordance with conventional detection methods for 241
Gram-positive bacteria and 95.7% for gram-negative bacteria (158). Yeast (C. albicans, C. parapsilosis, 242
12
C. tropicalis, C. glabrata and C. krusei) can also be identified using probes targeting 26S rRNA 243
sequences (74). Commercial methods give a result in 1.5 to 3 hours with high sensitivity and specificity, 244
but could not replace other detection methods as many significant pathogens are not detected using these 245
probes (table 2) (31; 87; 158; 242). 246
2. Microarray 247
Microarrays are based on species-specific or genus-specific DNA probes immobilized on chip on which 248
microorganisms DNA will specifically hybridize. The high-density of the chips allow the use of 249
sequences specific for different pathogens and probes for the detection of virulence factors and/or 250
resistance genes. Species identification is determined by the pattern of hybridization and on the intensity 251
of the signal (47; 179; 254; 264). The analytic sensitivity of microarrays ranges between 10 to 105 252
CFU/ml depending on the pathogen, which allows their use from positive BC but not directly from blood 253
(254). The Verigene Gram-Positive Blood Culture system and the Verigene gram-negative Blood Culture 254
system (Northbrook, IL, USA) can detect gram-negative or gram-negative microorganisms respectively, 255
as well as associated resistance genes. The time to identification is approximately 2.5 h from positive 256
blood culture bottles (table 2). After being extracted from 350 μl of blood culture from positive vial, the 257
bacterial DNA is suspended in a specific buffer and hybridized on specific synthetic oligonucleotides 258
followed by a second hybridization step involve gold particles. The signal is further amplified via a silver 259
staining process which increases the sensitivity of the system. The reading is automated. The Verigene 260
Gram-Positive Blood Culture system can detect 9 organisms at the species level (Staphylococcus aureus, 261
Staphylococcus epidermidis, Staphylococcus lugdunensis, Streptococcus anginosus group, Streptococcus 262
agalactiae, Streptococcus pneumoniae, Streptococcus pyogenes, Enterococcus faecalis and Enterococcus 263
faecium), 4 genus (Staphylococcus spp, Streptococcus spp., Micrococcus spp. and Listeria spp.) and 264
detect 3 resistance genes (mecA (methicillin), vanA (vancomycin) and vanB (vancomycin). Studies 265
performed on both adult and pediatric patients reported 89.7% to 99% of concordance for the detection 266
13
from positive BC between the Verigene Gram-Positive Blood Culture system and the traditional method 267
(5; 14; 158; 165; 206; 227; 258). Misidentifications have been reported for Streptococcus spp.. In 268
particular group mitis Streptococcus spp. have been misidentified with S. pneumoniae and S. oralis have 269
been misidentified with S. anginosus (206). Performances of the Verigene system are higher on mono-270
bacterial samples. Mixed culture can prevent the identification of one of the pathogen. Buchan et al 271
reported 98.6% of sensitivity and 94.3% of specificity for the detection of the mecA gene in 5999 272
Staphylococci cultures, and 100% of sensitivity and specificity for the detection the vanA gene in 81 273
cultures containing Enterococcus faecalis and Enterococcus faecium (26). Samuel et al reported 91% of 274
concordance for mecA and 100% for vanA detection when compared to routine methods (206). In case of 275
mixed population of staphylococci, the resistance gene cannot be associated to a specific pathogen (26). 276
The Verigene Gram-Negative can detect 5 pathogens at the species level (Escherichia coli, Klebsiella 277
pneumoniae, Klebsiella oxytoca, Pseudomonas aeruginosa and Serratia marcescens) and 4 genus 278
(Acinetobacter spp., Citrobacter spp., Enterobacter spp. and Proteus spp.). The Verigene Gram Negative 279
BC assay displays 91 to 100% of agreement for the detection of gram negative pathogens both from adult 280
and pediatric patients (17; 60; 108; 155; 226). Among, the false-negative results, mixed blood cultures 281
have been incriminated; but some of these negative or non called results involved bacteria covered by the 282
chip (60). As for many molecular assays, it is not possible to distinguish E. coli and Shigella spp. (S. 283
dysenteriae, S. flexneri, S. boydii, and S. sonnei). This distinction should be achieved with enzymatic 284
analysis. The Verigene Gram Negative assay can detect 6 resistance genes: the extended-spectrum beta-285
lactamase (ESBL) CTX-M and the carbapenemases KPC, NDM, VIM, IMP and OXA groups with 92.3% 286
to 100% of concordance with routine methods (60; 234). Other microarrays for the detection of resistance 287
gene are currently being developed (24). 288
The performances of microarrays are extremely high for both the detection of pathogen and resistance 289
cassettes when applied to positive blood cultures but supported by a limited number of studies. 290
14
3. Nucleic acid amplification-based methods 291
PCR provide a rapid and specific technology for pathogen identification from positive BC (260). The 292
specificity provided initially by the use of specific primers can be increased in the case of real time PCR, 293
by the use of specific probes. The use of PCR on blood or on hemorrhagic samples, was limited for long-294
time by the fact that PCR is sensitive to the presence of inhibitory compounds such as haemoglobin 295
contained in erythrocytes (2; 4; 111), lactoferrin contained in leukocytes (4), or immunoglobulin (3). To 296
avoid inhibition, recent methods use new or improved nucleic acid extraction and/or amplification 297
techniques. 298
Multiplex-PCR increases the time to result as it interrogates several targets (pathogens or resistance 299
genes) at the same time. In addition it can increase both the specificity and the sensitivity of an analysis 300
thanks to the use of multiple targets for the same organism (120; 223). Multiplexing is achieved through 301
the use of several specific primers pairs in the same reaction. Different methods can be applied to identify 302
the amplified sequence(s): 1) the use of specific probes labelled with distinct fluorochromes, 2) the 303
analysis of the probe or the amplicon melting curves (69; 72; 257); 3) the amplicon size (electrophoreses); 304
4) the sequencing of the amplification product. The latest technology, PCR/ESI-MS, is based on the 305
analysis of the amplicon by MS (Figure 5). The latter technique will be discussed in depth in the 306
paragraph on pathogen detection from blood. 307
Multiplex PCR 308
The FilmArray system (Idaho Technology, Salt Lake City, UT, USA) using multiple-PCR is a solution 309
that allows the identification of more than 25 pathogens and 4 antibiotic resistance genes from positive 310
BC in 1 h (19). The FilmArray is a closed system that uses multiple-PCR expected to identify 90 to 95% 311
of the pathogens involved in BSI as well as the resistance genes mecA, vanA and vanB and blaKPC . A kit 312
15
for the identification of potential bioterrorism agent (Bacillus anthracis, Francisella tularensis and 313
Yersinia pestis) is also available (249). 314
In a prospective study on 102 blood cultures, the FilmArray system displayed 91% of sensitivity and 77% 315
of specificity when compared to conventional identification methods for the detection of pathogen present 316
in the FilmArray panel only (19). The detection of resistance genes revealed 100% of specificity and 317
100% of sensitivity in this study. In another study, FilmArray displayed 98.5% of sensitivity and 100% of 318
specificity for the identification of gram-negative bacteria, 96.7% of sensitivity and 93.7% of specificity 319
form gram-negative bacteria and 96% of sensitivity and 98.9% of specificity for the detection of the mecA 320
gene. Another study performed on 118 mono-bacterial culture reported 92% of correct identification of 321
the FilmArray system, but mixed cultures gave no results (17). The identification of Candida albicans 322
and Candida glabrata reached 100% of sensitivity and 99.5% of specificity (6). This is consistent with 323
other study showing high performance of the FilmArray for the identification C. albicans and C. glabrata 324
(181). The FilmArray can also be applied on other sterile samples such as CSF (184). 325
Rapid PCR-based systems 326
The clinical and epidemiological impact of pathogens such as methicillin resistant Staphylococcus aureus 327
(MRSA) or rifampicin resistance Mycobacterium tuberculosis has contributed to the development of 328
methods for the rapid detection of resistance. S. aureus is a significant agent of community-acquired and 329
nosocomial infections (256). The mortality rate is increased in patient infected with MRSA in comparison 330
to patient infected with MSSA (methicillin sensitive S. aureus) (49; 50; 263). The increased percentage of 331
MRSA in comparison to MSSA has stressed the importance of rapid detections methods of MRSA (71; 332
77; 169; 170). Several methods based on multiplex real-time PCR are available: the GeneXpert 333
MRSA/SA BC Assay (Cepheid, Sunnyvale, CA) (45; 221), the StaphSR assay (BD GeneOhm, San 334
Diego, CA) (117; 222) and the StaphPlex (Genaco Biomedical Products, Huntsville, AL, USA) (230). 335
The GeneXpert and StaphSR assay are rapid PCR-based systems developed for the detection of S. aureus 336
16
only in approximately 1.5 h. They require only limited hands-on time and skills. The application of the 337
GeneXpert MRSA on positive blood culture was shown to have an impact on anti-infectious treatments 338
by reducing the use of glycopeptides (46). Alternatively, the StaphPlex system can identify several 339
staphylococci at the species level. The StaphPlex system is based on several PCRs (18 target genes) and 340
microarray analysis for the identification of staphylococci at the species-level, the detection of resistance 341
genes and of the Panton-Valentine leukocidin (PVL). The results are obtained within 5 hours (230). 342
Because these methods are largely aimed to detect resistance genes, they will be presented and discussed 343
in the chapter dedicated to resistance detection. 344
4. Matrix-Assisted Laser Desorption Ionization Time-Of-Flight Mass 345
Spectrometry (MALDI-TOF MS) 346
Principle of MALDI-TOF MS 347
The MALDI-TOF MS (Matrix-Assisted Laser Desorption Ionization Time-Of-Flight Mass Spectrometry) 348
is a method that allows identification of a broad range of microbes at the species or at the genus level 349
from positive culture. The method was first developed to be used on colonies before its application on 350
BC. Microbe identification using MALDI-TOF MS is based on the analysis of the mass spectrum 351
generated by bacterial or fungal component, mainly house-keeping proteins (the ribosomal proteins which 352
are basic are specifically extracted using acidic matrix). This spectrum is a unique fingerprint for each 353
microorganism but displays features shared between genetically related bacteria (44). The comparison of 354
the spectrum with a database of spectrum obtained for characterized organisms allow microbes 355
identification at the species or at the genus level depending of a score assigned by the identification 356
software (52). 357
MALDI-TOF MS was first used for bacterial identification in 1975 (8). The routine use of mass 358
spectrometry in diagnostic laboratory for the identification of bacteria was proposed in 1996 when correct 359
17
identification has been achieved directly from whole bacteria coming from single colonies (44; 110). 360
Identification of bacteria using the MALDI-TOF MS is divided in three steps: (1) ionization of the 361
sample, (2) separation of proteins in a flight tube (3) generation of a mass spectrum by determination of 362
the mass of the proteins (figure 3). The sample preparation consists in mixing the cells from a bacterial 363
colony, with a crystallizing matrix that will trigger the ionization of the sample achieved by a laser 364
(MALDI). Protein accessibility can be facilitated by a quick formic acid extraction. Ionized proteins are 365
accelerated by an electrostatic field and separated along a flight tube. The separation is due to the time of 366
flight (TOF), a function of the mass and of the charge of the proteins. Protein detections will generate a 367
mass spectrum unique for a defined bacterial strain corresponding to mass-to-charge ratio (m/z) between 368
1000 and 20000 kDa. However, several peaks are shared between bacteria from the same genus or species 369
and serve as biomarkers for bacterial identification. The identification is performed by comparison of the 370
mass spectrum with a database of spectra obtained from characterized bacterial strains. The software that 371
performed the spectral analysis and the comparison with the database, assign a score for the identification. 372
Depending on the threshold recommended by the furnisher and the in-house algorithm, the identification 373
can be rejected, or accepted either at the species level or only at the genus level. 374
Application of MALDI-TOF MS on positive blood culture bottle 375
In clinical microbiology, the MALDI-TOF MS was first used on pure bacterial colonies before its 376
application on positive BC for the diagnostic of BSI (figure 3) (52). The initial analysis on a positive 377
blood culture bottle is a Gram staining. The Gram staining is especially recommended 1) to confirm that 378
the detection is truly due to the presence of a microorganism and 2) give a presumptive etiology of the 379
BSI and 3) to disclose mixed bacteremia. In positive blood cultures, the bacterial concentration is ranging 380
from 106 to 109 CFU/ml, which in theory is sufficient for MALDI-TOF MS identification. However, the 381
presence of high amount of non bacterial material (erythrocytes, nutrient from the growth media) impairs 382
the direct identification from the blood vial. 383
18
Application of MALDI-TOF MS after a short subculture 384
A short subculture (2 to 3 hours) on agar plates can be performed from the positive blood vial to obtain a 385
thin layer of bacteria (174). From this thin layer, the MALDI-TOF MS analysis can be performed (figure 386
2) (116; 239). Short subcultures are also adequate for AST using automated systems (116). However, this 387
procedure is mainly suitable for fast growing bacteria. 388
Application of MALDI-TOF MS directly from the blood culture vial via a bacterial pellet 389
The fasted method is to perform the identification directly from the positive BC (figure 2 and figure 3). 390
The goal is to get rid of non bacterial components present in the bottles such as red blood cells and to 391
concentrate the bacteria. Bacterial concentration can be achieved by centrifugation and erythrocyte lysis 392
by ammonium chloride (51; 193). This generates a pellet that can be identified by MALDI-TOF MS 393
(figure 3). This method allowed the identification of 78.7% of the samples obtained. Moreover, 99% of 394
the MALDI-TOF MS identifications were correct at the species level (193). Alternative erythrocytes lysis 395
techniques can be achieved with formic acid (136). Any methods implying mild detergent that would 396
solubilise erythrocytes membrane but not microbe membrane lead to similar results (83; 163; 168). 397
Alternatively, gel-based separator tubes have proven efficiency in concentrating bacteria and removing 398
red blood cells (224). 399
MALDI-TOF MS has a major impact on the time to result since it can be achieved directly from positive 400
blood culture, without subculture (61). The sample processing and identification using MALDI-TOF MS 401
takes approximately 1 hour, which makes it time effective for fast growing bacteria as well (136; 193). 402
The accuracy of the MALDI-TOF MS identification is dependent on extended and correct spectrum 403
database. This database can be implemented with spectrum obtained from clinical isolates. This is 404
applicable for strains for which the MS identification failed and that were identified without any doubt 405
with other methods such as gene sequencing or enzymatic assay. This is also applicable for rare 406
19
pathogens that can be identified faster and accurately using MALDI-TOF MS (176; 213). However, some 407
closely related bacterial species such as the different Streptococcus species remain difficult to distinguish 408
using MALDI-TOF MS. For instance, group mitis Streptococcus are often misidentified as S. 409
pneumoniae. S. pneumoniae identification should thus be confirmed by phenotypic test such as optochine 410
susceptibility (126; 224; 227). Similarly, the distinction of E. coli and Shigella sp. is impossible using 411
routine MALDI-TOF MS procedures and requires additional phenotypic confirmation (59; 210). 412
Recently, a new approach based on the analysis of biomarker peaks has been proposed to differentiate 413
these two closely related pathogens (128). 414
Efficiency and reproducibility of MALDI-TOF MS identification 415
Even if a mass spectrum can be obtained from a single colony, this biomass represents a significant 416
amount of bacteria. MALDI-TOF MS identification is therefore not applicable directly from blood but is 417
dependent on microorganisms proliferation in BC and on an additional concentration step from the 418
positive BC (174). The requirement of a culture limits the use of the MALDI-TOF for the detection of 419
non-cultivable bacteria. 420
Proteins used as biomarkers for MALDI-TOF MS identification of bacteria are mainly involved in house-421
keeping functions (246). Ribosomal proteins contribute to approximately half of the peaks present in the 422
mass spectrum (10; 201). This makes it reproducible and robust as several peaks would be conserved in 423
genetically related bacteria. Nevertheless, experimental conditions can impact the presence and the 424
relative abundance of detected peaks. This is true for both routine identification or for the implementation 425
of the peaks pattern into the in-house database. The culture media, the growth conditions and the age of 426
the colony may impact the spectrum (235). The presence of agar residue or blood proteins in the sample 427
may impact the spectrum quality (52). Quality of the matrix may also influence the spectrum (246). 428
20
Importance of the Gram staining and of the subculture 429
The Gram staining on positive blood culture is still mandatory (i) to confirm the presence of a microbe in 430
the bottles (ii) to validate the identification of the MALDI-TOF MS, which should be congruent with the 431
characteristic provided by the Gram (88) (iii) to disclose polymicrobial infections as the MALDI-TOF 432
MS poorly identify mixed bacterial samples. Mixed infection may also be detected upon subculture, 433
which also remains essential for the antimicrobial susceptibility testing (see later). 434
435
MOLECULAR METHODS FOR THE DETECTION OF PATHOGENS 436
DIRECTLY FROM BLOOD 437
BSI is initially suspected by clinicians on the basis of clinical signs and symptoms. At this stage the 438
etiological agent is however difficult to suspect as the clinical presentation is generally similar. To 439
accelerate the time to result of microbial diagnosis, molecular methods that can be used directly on whole 440
blood have been proposed. Nevertheless, blood as sample presented many technical limitations: 1) the 441
low quantity of circulating microbes during BSI (1 to 10 CFU per ml), 2) the presence of PCR inhibitor in 442
blood that are not completely removed by current nucleic acid extraction methods (see Introduction 443
section), 3) the sensitivity of PCR to contamination; which implies the use of highly pure (nucleic acid 444
free) material and ideally to perform quantitative analyses (211), 4) the ability of PCRs to detect the 445
presence of DNA from both living and dead microbes which makes the interpretation of positive results 446
difficult (BSI versus DNAemia and 5) the presence of human DNA in excess . 447
Different PCR technologies are now available. Broad-range PCRs could be useful, especially when the 448
etiological agent cannot be suspected, but they are often limited by lower sensitivity. Pathogen specific 449
PCRs display higher sensitivity. Real-time PCRs can be multiplexed to detect several targets at the same 450
21
time with a good sensitivity. A recent innovation that can also be use directly on whole blood is PCR-451
ESI-MS that is presented as a universal and fast method. 452
1. Broad range PCR 453
The need for a PCR-based universal detection method is stressed by the importance of rapid diagnostic in 454
the setting of critically-ill patients, especially since the etiology of the BSI is difficult to establish on the 455
basis of the clinical presentations. This is the case for neonate patients with the additional difficulty that 456
only small volumes of blood are available, which decreases significantly the sensitivity of BC. 457
Alternatively, such universal and sensitive methods may be used to screen blood samples taken from 458
neutropenic patients or other high risk subjects to early detect the etiological agent of a BSI. 459
The 16S rRNA gene that displays sufficient level of conservation among bacterial species is the target of 460
in-house PCRs used in many molecular diagnostic laboratories (12; 40; 41; 43; 94). In many organisms, 461
the ribosomal operon is present in multiple copies which increases the sensitivity of PCRs targeting this 462
genomic region. The intergenic spacer is more polymorphic and more species specific. Specific PCR 463
primers are designed on conserved regions that surround highly variable region. Thus, the sequencing of 464
the amplicon allows the identification of the microbe at the genius or at species level. Most of these PCR 465
cannot be used directly on blood because of their limited sensitivity and specificity. However, some 466
studies have reported sensitivity of PCR targeting the 16S rRNA gene ranging from 10 to 2.5.102 CFU 467
per reaction directly from blood (131; 191). Gaibani et al have developed a broad-range real-time PCR 468
that targets a 97 base pair sequence of the 23S rRNA gene. This PCR is expected to detect 90% of the 469
bacteria involved in BSI but does not give any identification, since the short fragments amplified do not 470
allow discrimination upon sequencing (89). This real-time PCR can be used on whole blood (100 μl). Its 471
sensitivity is ranging from 10 to 103 CFU per reactions for E. coli and S. aureus, depending on the 472
extraction method. 473
22
The SepsiTest system (Molzym, Bremen, Germany) is based on broad-range PCR using universal primers 474
that target the 16S rRNA gene for bacterial identification and the 18S rRNA for fungal identification. The 475
amplicon are sequenced and analysed by BLAST to give identification at the species level (=> 99% of 476
identity) or genus level (=> 97%) (134; 250). Of course this cut-off does not apply to all bacterial species 477
and should be used with caution. The analysis is achieved from 1ml of whole blood. After lysis of human 478
cells, the human DNA is degraded by a DNAse. The PCR is achieved using reagent provided by the 479
manufacturer and lead to amplicon of about 450 base pairs. The time to result is approximately 6 hours. 480
Depending on the study, the sensitivity and specificity are variable when compared to BC. A first study 481
performed in critically-ill patients led to 28.6% of sensitivity and 85.3% of specificity of the SepsiTest 482
when BC is used as gold standard (145; 212). In a second study, pathogen could be detected in 26 % 483
(13/50) of the critically-ill patients while SepsiTest could detect pathogens in only 12 % of the patients 484
(212). In a third study, still using BC as gold standard, SepsiTest sensitivity was 21% and specificity was 485
96% (152). In contrast, high sensitivity and acceptable specificity (87% and 85.5%) was monitored in a 486
multicenter study involving 342 blood samples from 187 patients (250). In a study performed on patients 487
supported by extracorporeal membrane oxygenation, the sensitivity and specificity were 78.6 % and 88.4 488
% when compared to BC with 97.7% of concordance of positive results (177; 200). In this study, 489
SepsiTest could detect at least a pathogen in 25% of patients with negative BC. For patient with suspected 490
endocarditis, Kuhn et al. reported higher performances of SepsiTest over BC (134). The variable 491
sensitivity and specificity of the SepsiTest when compared to BC needs to be investigated. One specific 492
feature of SepsiTest is that it can be used on sterile samples other than blood with the potential to replace 493
home-made broad-range PCR (96; 200). A limitation of this method is that the extraction is not automated 494
(99). Another important limitation of all these molecular tests is that they do not provide strain for 495
detailed antibiotic susceptibility testing. 496
The VYOO system (SIRS-Lab, Jena, Germany) is based on multiplex PCRs that can detect 34 bacteria 497
and 7 fungi from 5 ml of whole blood. The system contains several steps, including non-automated steps. 498
23
Nucleic acid extraction is assisted by magnetic beads. Then there is a step of enrichment in microbial 499
DNA based on i) the methylation difference between microbes (bacteria and fungi) DNA and the human 500
DNA and ii) the use of chromatography affinity (202). The PCR is achieved on DNA which concentration 501
has to be manually adjusted to 1 μg in 25ul. Two PCRs with a specific pool of primers are required. The 502
amplicons are applied to an electrophoresis in an ethidium bromide stained agarose gel. Pathogen 503
identification is determined on the basis of the band pattern. 504
In a study performed on 311 blood samples from 245 patients of an ICU of a university hospital analysed 505
in parallel with BC and with the VYOO system, the VYOO system gave 30.1% of positive samples, with 506
a median time to result of 24.2 hours and BC bottle gave 14.5% of positive samples with a median time to 507
positivity of 68 hours with only 40% of correlation between the two methods (22). PCR positive samples 508
correlated well with the level of procalcitonin. Additionally the PCR results from blood were most of the 509
time confirmed by the pathogen identified in the suspected site of infection. Importantly 34% of the 510
patients with positive VYOO test, the anti-infectious therapy was inadequate, including infections that 511
involved VRE, multi-drug resistant staphylococci and fungi (22). In another study, the VYOO gave 512
positive results in 51.4% of samples from patient with BSI, when BC gave 27.7% of positive results (84). 513
These studies suggest that VYOO is more sensitive than BC, which could be explained by the association 514
of a large starting volume (5 ml) and the enrichment in microbial DNA before amplification. More studies 515
are required but these preliminary performances are promising despite the fact that many steps are not yet 516
automated and that. 517
2. Real-time PCR and Multiplex PCR 518
519
The multiplexing of real-time PCRs allows detecting simultaneously several microbes with a good 520
sensitivity and specificity. This is the technology used for the LightCycler SeptiFast system (Roche 521
Molecular System, Germany), the MagicPlex Sepsis system (Seegene, Seoul, Korea). 522
24
Principle of the LightCycler SeptiFast system 523
The LightCycler SeptiFast system is based on broad-range real-time PCR that can identify 19 pathogens 524
(8 gram-negative bacteria, 6 gram-negative bacteria and 5 fungi) representing approximately 90% of the 525
pathogen responsible for BSI. The SeptiFast can be used directly from whole blood with an overall time 526
to result of about 5 hours (3.5 when associated to an automated extraction system). The targets of the 527
multiplexed broad-range real-time PCR are the internal transcribed spacer (ITS) between the 16S and the 528
23S ribosomal DNA for bacteria and the ITS between the 18S and the 5.8S ribosomal DNA for fungi. 529
Identification at the species level is determined by the distinct melting curves of the specific probes. The 530
LightCycler can detect 8 gram-negative bacteria detected: E. coli, Klebsiella pneumonia/ Klebsiella 531
oxytoca, Enterobacter cloacae/ Enterobacter aerogenes species, Proteus mirabilis, Pseudomonas 532
aeruginosa, Acinetobacter baumannii, Stenotrophomonas maltophilia. Six Gram-positive bacteria can be 533
recognized: S. aureus, coagulase negative staphylococci, Streptococcus pneumoniae, Streptococcus 534
pyogenes, Streptococcus agalactiae / Streptococcus mitis, Enterococcus faecium and Enterococcus 535
faecalis. Six fungi can be detected and identified at the species level: Candida albicans, Candida 536
tropicalis, Candida parapsilosis, Candida glabrata, Candida krusei and Aspergillus fumigatus. Basically, 537
the assay consists of three separated reactions that contain distinct primers and probes mixes to detect 538
respectively gram-negative bacteria, gram-negative bacteria and fungi. The LightCycler SeptiFast is 539
designed for the analysis of 1.5 ml of blood on which a mechanic cell lysis is achieved with the SeptiFast 540
Lys KIT MGRADE and the MagNALyser from Roche diagnostic prior DNA extraction using also a 541
commercial kit provided by the same manufacturer. The assay contains an internal control that consist of 542
a synthetic double-strand DNA, similar to the expected amplicon but with a distinct probe binding site. 543
The analytic sensitivity is ranging from 3 to 30 CFU/ml for bacteria and is 100 CFU/ml for fungi (34; 544
252). The experimental sensitivity (42.9% to 95%) and specificity (60% to 100%) of the SeptiFast are 545
variable depending on the studies and on the patients characteristics (table 1) (21; 35). 546
25
Diagnosis of Infection in febrile neutropenia using the LightCycler SeptiFast 547
From a group of 86 febrile neutropenic patients representing 141 episodes of fever, BC and SeptiFast 548
detected approximately the same number of microorganisms: 44/141 (31.2 %) and 42/141 (29.8%), 549
respectively. However, the association of BC and SeptiFast increases the rate of documentation from 550
about 30% to 43% (61/141) which might be due to the fact that only 12 organisms where detected by both 551
BC and SeptiFast (139). A similar observation was made by Mancini et al. on a study performed on 103 552
samples from neutropenic patient with haematological diseases in which 20.4% of the samples were 553
positive with BC and 33% using SeptiFast, with only 83% of correlation between the two methods (154). 554
For Bravo et al., the agreement between BC and SeptiFast was 69% for neutropenic patients and 75% for 555
patients from the intensive care unit (25). All these studies suggest that SeptiFast cannot replace BC but 556
that the two techniques could be complementary (98; 139; 154; 161). Lamoth and colleagues suggested 557
that the low sensitivity of SeptiFast was due i) to organisms absent from the SeptiFast analytic spectrum 558
(40% of false negative) and ii) to the cut-off that decrease the sensitivity for the detection of coagulase 559
negative staphylococci or for streptococci and increased the rate of false negative (139). False negative 560
results obtained for gram-negative cocci might also be explained by the inefficient lysis of these 561
microorganisms (182) or to PCR inhibition in the case of high bacterial load. The adjustment of the 562
experimental procedure and an adequate cut-off might increase the performance of the SeptiFast assay for 563
neutropenic patients (139; 182). In a context of persistent fever, SeptiFast identified new pathogens in 564
89% of the cases whereas BC identifies 8% of the pathogens (139). Similar results were obtains by von 565
Lilienfeld-Toal et al. (241). In this study all the patients with a probable fungal infection had positive 566
SeptiFast results for Aspergillus fumigatus. This suggests a potential added value of the SeptiFast assay 567
for the detection of fungemia in neutropenic patients (35; 139; 154; 180; 241). 568
26
Diagnosis of infectious endocarditis 569
A study performed on 63 patients with suspected endocarditis revealed a low sensitivity of the SeptiFast 570
when compared to BC (34). Among 19 patients with positive blood culture at their admission, SeptiFast 571
detected 8 bacteria (41% of sensitivity). Twenty-two patients had positive blood culture before their 572
admission but because of efficient antibiotic treatment they had negative blood culture at the time of this 573
study. Among them, SeptiFast detected only 3 bacteria (3/22) (34). In this study, SeptiFast did not detect 574
any microbes for patients without any etiology (100% of specificity) (34). Because the low sensitivity of 575
the SeptiFast was not due to pathogens that are not detected by this system, it could be due as discussed 576
later to a low bacterial load. In another study involving 20 patients with endocarditis, SeptiFast performed 577
on excised cardiac valves, displayed a higher sensitivity (95%), and specificity (100%) than culture 578
(sensitivity 15% and specificity 100%) (81; 147). In patient with suspicion of Candida spp. endocarditis, 579
SeptiFast was as sensitive as BC to detect fungal infection (141). 580
Diagnosis of neonatal sepsis 581
In a study performed on 1,673 pediatric samples (803 children), the detection rate of infection was higher 582
with the SeptiFast (14.6%) than with BC (10.3%) which corresponded to a sensitivity of 85.0% and a 583
specificity of 93.5% of the SeptiFast (153). The cumulative positive rate of BC and SeptiFast was 16.5%. 584
Another study confirmed a higher sensitivity but a lower specificity of SeptiFast for the detection of late-585
onset neonatal sepsis (127). 586
Advantages and inconvenient of the LightCycler SeptiFast system 587
The performances of the SeptiFast are variable depending on the studies. The overall sensitivity of the 588
SeptiFast is not higher than the sensitivity of BC but together SeptiFast and blood culture increase the rate 589
of microbial documentation of BSI. This suggests that the SeptiFast cannot replace blood cultures but that 590
these two methods are complementary. One of the major advantage of the SeptiFast is the time to result 591
27
(<5hours). In a study performed on 114 consecutive patients with clinical evidence of sepsis the mean 592
time to results for SeptiFast was less than 8 hours when BC mean time to positive result was 3.5 days and 593
for negative result 5 days (218). In a study performed on patients of the emergency department with 594
suspected sepsis, Schaub et al. Reported a median time to positivity of BC of 16 hours (without the 595
organism identification) (209). SeptiFast might be useful in case of a persistent bacteremia, for instance in 596
neutropenic patient with persistent fever, for which SeptiFast often identified additional pathogens. There 597
are a high number of studies on the performance of the SeptiFast. However, there is a need of 598
interventional studies to clearly determine the clinical impact of this new technology. 599
One of the limits of the SeptiFast is also the absence of exact quantification. Cut-off can be used for the 600
interpretation of streptococci and coagulase negative staphylococci positive results in particular. This is 601
expected to reduce the number of false positive results but might prevent the detection of low grade 602
infections. In some studies performed in neutropenic patients, the sensitivity of SeptiFast for the detection 603
of coagulase negative streptococci is decreased, which could be due to the fact that the cut-off were 604
established for non-neutropenic patients. In contrast in pediatric studies SeptiFast leads to increasing 605
number of false positive due to coagulase streptococci. Adding quantitative data to the SeptiFast has the 606
potential to predict the severity of the sepsis (266). 607
The HACEK group of fastidious bacteria (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella and 608
Kingella) involved in a significant number of blood culture-negative bacteremia is not covered by the 609
SeptiFast. For this reason SeptiFast is not sufficient by itself to diagnose all the bacteremia and alternative 610
methods will still be required to cover the entire bacterial kingdom. Finally, in the context of an 611
increasing number of multi-drug resistance organisms, it is a limitation that SeptiFast does not provide 612
any data on the resistance profile of identified organisms excepted for MRSA but using an additional kit 613
(table 1). 614
28
The MagicPlex Sepsis system 615
The MagicPlex Sepsis system (Seegene, Seoul, Korea) is based on multiplexed real-time PCRs that can 616
detect 90 pathogens at the genus level, 25 at the species level (19 bacteria and 6 fungi), and the resistance 617
genes mecA, vanA and vanB directly from whole blood (1 ml). A specific nucleic acid extraction kit is 618
used to enrich in microbial DNA. A first amplification is a screening that provides either (i) an amplicon 619
bank that consists of gram-negative bacteria and fungi or (ii) an amplicon bank that consists of gram-620
negative bacteria (n=73) and resistance genes conferring resistance to methicillin (mecA) or vancomycin 621
(vanA and vanB) . A second step is a screening for bacteria and fungi at the genus level and for the 622
presence of resistance genes (time to result 5 hours). 623
The performance of the MagicPlex system was compared to BC on 267 patients from the intensive care-624
unit, from the haematology department and from the emergency department of a tertiary hospital, which 625
revealed an agreement between the two methods of 73 % with no statistical difference between their 626
sensitivity (30). For patient with antibiotic treatment, the sensitivity of the MagicPlex 65% was lower 627
than the sensitivity of BC (71%); specificity was respectively 92% and 88% for each method. Another 628
study performed on 140 patients with suspected BSI reported 37% of sensitivity and 77% of specificity of 629
the MagicPlex when considering BC as gold standard (152). 630
Additional studies are required to determine the exact potential of this device. This method is limited by 631
the number of pathogen detected at the species level and by the absence of quantification. Another 632
limitation is the absence of automation and the need of a specific extraction method which might makes it 633
difficult to integrate in an automated molecular diagnostic laboratory. 634
635
636
29
3. PCR-Electrospray Ionization Mass Spectrometry (PCR/ESI-MS) 637
Principle, characteristics and performance of PCR/ESI-MS 638
PCR-Electrospray Ionization Mass Spectrometry (PCR/ESI-MS) is a technology that was initially 639
developed to face bioterrorism threats or to run public health investigations. The aim was to provide a 640
rapid method for the detection and identification of pathogens from various type of sample. The 641
technology had to detect low quantity of a given pathogen even in polymicrobial samples and should 642
include the detection of non cultivable or fastidious microorganisms. In addition, the technology should 643
be able to detect known microorganisms as well as yet unknown pathogen. Molecular diagnosis was 644
chosen as it is fast and broad range. 645
The PCR/ESI-MS approach is based (i) on the amplification of microorganisms DNA by multiple PCRs 646
and (ii) on the identification of the organism(s) at the species or genus level through the analysis of the 647
amplicon by ESI-MS (Electrospray Ionization Mass Spectrometry) (65; 123). Basically, PCR/ESI-MS 648
consists in 5 steps: 1) extraction of the microorganism or sample DNA, 2) amplification of the DNA 649
using multiple pairs of primers, 3) precise determination of the molecular mass of the amplicon(s) using 650
ESI-MS, 4) deduction of the base composition of the amplicon(s) from the exact MW of the amplicon 5) 651
identification of the pathogen(s) by integrating the informations obtained from several amplicon(s) (figure 652
4) (66; 123). 653
Before being applied to blood (see next paragraph) PCR/ESI-MS was first applied on bacterial colonies 654
and on environmental samples. Practically, DNA amplification is based on multiple broad-range PCRs, 655
which is more sensitive, than a single broad-range PCR based on degenerated primers. The primers are 656
targeting conserved genomic region surrounding polymorphic regions (for instance DNA regions 657
encoding for the 16S or the 23S rRNA). Moreover, these PCRs are devoid of fluorescent dye or probes, 658
which allows high-level of multiplexing. DNA amplification may be performed in 96 well plate using the 659
30
following steps: 1) 95°C for 10 min, 2) 8 cycles of 95°C for 30 s, 48°C increasing of 0.9°C at each cycles 660
for 30 s and 72°C for 30 s 3) 37 cycles of 95°C for 15 s, 56°C for 20 s, and 72°C for 20 s 4) final 661
extension of 2 min at 72°C and hold at 4°C hold (73). Major innovation of PCR/ESI-MS is that the 662
pathogen’s identification relies only on the base composition (A, C, G and T) of the amplicons (65; 123). 663
The base composition is obtained by the integration of the exact mass of the amplicon, the length of the 664
amplicon, the mass of each base and the complementarily rules of DNA (figure 4). The choice of the 665
PCR targets makes the base composition of one or more amplicon(s) sufficient for pathogen 666
identification. The absence of Sanger sequencing dramatically reduces the time to result(s). Thus, ESI-MS 667
provides the base composition of an amplicon in about 1 minute, the mass of the amplicon being 668
determined by the time of flight (TOF). The base composition of the amplicon(s) is compared to a 669
database providing the identification at the genus or at the species level with a score of probability (Q-670
score) inferred by an ESI-MS triangulation software relying on multiple amplified regions. The Q-score 671
integrates multiples parameters such as the number of primer pairs that gave an amplicon, the number of 672
potential microorganism, the proximity of the base compositions to reference matches in the database. 673
A set of 9 pairs of primers is used for the coverage of the bacteria kingdom. Four pairs are necessary for 674
Candida species. Four additional pairs of primers have been designed to detect the resistance cassettes 675
mecA, blaKPC, vanA and vanB. Using clinical samples (up to 1.25 ml) other than blood, PCR/ESI-MS 676
displays good sensitivity and specificity. This includes environmental samples as well as clinical samples 677
such as CSF (75; 76; 171) and respiratory tract samples (64; 91; 119). Compared to culture the PCR/ESI-678
MS correctly identified 95.6% and 81.3% of the strains at genus level and species levels respectively. 679
Among 395 respiratory samples, PCR/ESI-MS displayed 67.6% of agreement at the genus level and 680
66.6% at the species level with culture. PCR/ESI-MS was able to identify fungi in 20.3% (35/172) of the 681
respiratory specimens with a negative culture suggesting a better sensitivity than culture (217). A 682
commercial kit available for the detection of viruses (38; 39; 48; 107; 135; 142; 185; 228; 231) has been 683
shown to display higher performance than real-time PCR and microarrays (228). PCR/ESI-MS have also 684
31
been successful at identifying Mycobacterium species at the species level and at determining associated 685
resistance genes using a panel of 8 PCRs (138; 160; 245). 686
Thus, PCR/ESI-MS represents a universal method that may be applied to bacteria, viruses and fungi and 687
that is expected to also identify any unknown species. Indeed, when primers are designed to identify all 688
known members of characterized groups, they allow the detection of unknown or new emerging 689
pathogens from these groups (138; 203; 204). This provides a potential for the rapid detection of 690
emerging pathogens. In addition, PCR/ESI-MS can detect and identify all different pathogens present in 691
polymicrobial samples with quantitative results. This method can also be used on formalin or paraffin 692
embedded tissue (216). 693
Finally, PCR/ESI-MS has been extensively used for genotyping and serotyping of bacteria and viruses 694
because of its accuracy to detect single nucleotides variations, which represents a tool for health care 695
epidemiological investigation or outbreak follow-up.(23; 62-64; 67; 101; 102; 159; 190; 207; 214; 245; 696
262). Such an accuracy of the PCR/ESI-MS naturally relies on a representative database and on its 697
maintenance. 698
Application of PCR/ESI-MS to the diagnosis of BSI directly from blood 699
The first instrument that was developed was the TIGER (Triangulation Identification for the Genetic 700
Evaluation of Risk) that allowed the detection of specific organisms even in polymicrobial samples (204). 701
The Ibis T5000 instrument, the first commercial version of TIGER could identify up to 800 pathogens 702
from whole blood (65; 261). When compared to blood culture, PCR/ESI-MS applied on 1 ml of whole 703
blood displayed 50% of efficiency. A second commercial version, namely the PLEX-ID (Abbott), using 704
up to 1.25 ml of blood, has been developed displaying good sensitivity and specificity on most of the 705
samples which includes environmental samples as well as clinical samples such as CSF and sputum as 706
said above (15; 38; 76; 91; 138). The first procedure of the PLEX-ID for whole blood consisted of a 707
mechanic cell lysis, using beads, facilitated by the addition of proteinase K, SDS and eating at 56°C. 708
32
DNA purification was then achieved with magnetic silica beads and eluted in 250 μl of water at 70°C 709
(137). The PCR is based on 9 pairs of primers for bacteria detection, 4 pairs for Candida species and 4 710
pairs for detection of resistance cassettes (mecA, blaKPC, vanA, vanB). One additional pair of primer 711
corresponds to the extraction control. The PLEX-ID gave 78.6% of agreement with blood culture over 712
906 specimens taken from adult and pediatric patients (464 with positive blood culture and 442 with 713
negative blood cultures). While 33 culture negative specimens were PLEX-ID positive, 97 culture 714
positive specimens were PLEX-ID negative (137). In this study the estimated sensitivities of PCR/ESI-715
MS was 85.9% when the estimated sensitivity of BC was 41.2%. To further increase the sensitivity of the 716
detection directly from whole blood a second version of the PLEX-ID that could analyse up to 5 ml of 717
blood has been developed. In this version, the increased sample volume was associated with a DNA 718
extraction method and PCR conditions optimized for whole blood samples (13). Practically, 5 ml of 719
whole blood were lysed in presence of 665 μl of a commercial buffer (100 mM Tris solution containing 720
guanidinium thiocyanate and detergent), 145 μl of BSA 10% containing a pumpkin DNA extraction 721
control), and yttria-stabilized zirconium oxide beads (166). After the removal of lysed red-blood cells by 722
centrifugation the nucleic acid were extracted from the supernatant using silica-coated magnetic beads. 723
Thirty microliters of the eluate were used for the PCR with 25 μl of PCR master mix containing the pool 724
of primers pairs (137). Rather than trying to remove the excess of human DNA, PCR conditions were 725
defined in the context of high concentration of human DNA. This was achieved by testing multiple PCR 726
conditions by modulating Mg2+, primers and polymerase concentration and annealing temperature in the 727
presence of 12 μg of human DNA (13; 73). In this background, high primers concentration (750 μM) 728
together with high polymerase concentration (2.2 units per reaction) resulted in a PCR yield of 86% of the 729
yield when 1 μg human DNA was present. Modulating Mg2+ concentration or the annealing temperature 730
had only a weak impact on the PCR yield. This procedure was then evaluated in a prospective study 731
involving 331 patients with suspected BSI (13). For each patient 2 blood bottles (one aerobic and one 732
anaerobic) were inoculated with 5 ml of whole blood and an additional blood sample was collected for 733
PCR/ESI-MS analysis from the same venipuncture (137). The PCR/ESI-MS displayed 83% of sensitivity 734
33
and 94% of specificity as compared to BC. Interestingly, this corresponded to 35 positive specimens 735
(10.6%) by PCR/ESI-MS and 18 positive by culture (5.4%). In the absence of any method to investigate 736
the discrepant results, a second aliquot was analysed using PCR/ESI-MS, which confirmed almost all the 737
identifications and increased the sensitivity of PCR/ESI-MS to 91% and the specificity to 99%, were (13). 738
The limit of detection of the PCR/ESI-MS method developed by Bacconi et al. is 16 CFU per mL for S. 739
aureus, K. pneumoniae and E. faecium, and 4 CFU per mL for C. albicans. It is generally admitted, based 740
on plating methods, that the number of bacteria circulating during BSI is between 1 and 10 CFU per ml 741
(105; 132; 133; 243; 251). The analysis of the data from the literature performed by Bacconi and 742
colleagues estimated that the amount of bacterial DNA in blood during a BSI vary between 103 to 104 743
genomes copies per ml (13). This difference could be due to the fact that plating methods reflect the 744
number of cells that survive the procedure rather than the number of circulating cells. Molecular diagnosis 745
directly from blood could detect also free DNA, DNA resulting from dead bacterial cells and bacterial 746
DNA present within phagocytes, which may explain the high performance of the PLEX-ID on whole 747
blood. However, this extremely high sensitivity may lead to contamination by DNA. For this reason, the 748
use of ultra-clean reagents associated with molecular biology laboratory practices is mandatory. 749
PCR/ESI-MS is also a promising tool for the diagnostic of blood culture negative BSI. Indeed, among 750
464 patients with positive blood cultures and 442 patients with negative blood culture, 33 culture negative 751
cases were detected positive by PCR/ESI-MS and these cases were demonstrated to be true BSI based on 752
the analysis of the clinical presentation (137). Using primers and probe specific of Ehrlichia chaffeensis 753
targeting the 16S gene and a genus specific set of primers and probes Eshoo and colleagues reported 754
18.8% of PCR/ESI-MS positive specimen of whole blood among 213 blood samples from patients with 755
suspicion of ehrlichiosis (73). From the same pool of specimen, PCR/ESI-MS identified Rickettsia 756
rickettsii from 4 samples and Neisseria meningitidis from one samples. 757
PCR/ESI-MS was designed to detect all pathogens present in a mixed microorganism populations (118; 758
205). This is also true for polymicrobial blood culture since PCR/ESI-MS could identify 29 bottles with 759
34
mixed populations out of 234 positive BC bottles (125). A quantitative analysis performed using an 760
internal control helps at (i) interpreting results for polymicrobial samples and (ii) identifying true positive 761
versus contaminating organisms (137). Further studies are required to determine if the quantification 762
could also help at defining the severity of the infection. 763
In conclusion, the PLEX-ID, which provides reliable results in less than 6 hours, is a versatile system that 764
may be directly used on blood starting from 5 ml of samples and that represents a complementary 765
approach to blood cultures. 766
MOLECULAR DETECTION OF ANTIMICROBIAL RESISTANCE 767
DURING BLOOD STREAM INFECTIONS 768
BSI is characterised by a high level of morbidity and mortality that increases with the delay in the 769
introduction of an efficient anti-infectious therapy (58; 82; 90; 113; 148; 149; 157; 232). The rapid 770
identification of the pathogen gives a first indication on the effectiveness of the empiric therapy. 771
However, rapid information on the resistance profile of the agent of a BSI, would allow to better adjust 772
the empirical treatment. 773
1. Rapid PCR Based-methods for resistance detection from positive blood 774
cultures 775
Detection of resistance mechanisms of gram-positive bacteria 776
Staphylococcus aureus is a significant agent of both community-acquired infections and nosocomial 777
infections (256). In recent years, it has been observed an alarming increase of the percentage of 778
methicillin resistant Staphylococcus aureus (MRSA) a drug resistant pathogen associated with significant 779
increase of the mortality (49; 50; 169; 170; 263), which explains that many rapid methods are now 780
35
available to simultaneously identify S. aureus and detect the mecA gene associated with methicillin 781
resistant (table 1) (71; 77). 782
The GeneXpert MRSA/SA BC Assay (Cepheid, Sunnyvale, CA) is a good example of such dual detection 783
method based on real-time PCR for the identification of S. aureus and the detection of the mecA gene. 784
The analysis may be performed directly from positive BC in about 1 hour and was validated both in adult 785
and pediatric patients with BC positive for gram-negative cocci in cluster with 100% of sensitive and 99.5 786
to 100 % of specificity (45; 221). The GeneXpert MRSA/SA BC Assay dramatically reduced the time to 787
detection of S. aureus and has an impact on the detection of MRSA (figure 2) (45; 54; 140; 208). 788
Also based on real-time PCR, another rapid PCR-based assay, the StaphSR assay (BD GeneOhm, San 789
Diego, CA), can identify and differentiate methicillin-susceptible S. aureus (MSSA) and methicillin-790
resistant S. aureus (MRSA) from positive blood cultures in about 1.5 h with an analytical sensitivity of 15 791
DNA copies per reaction mixture which correspond to 103 DNA copies per ml (117; 222). The assay was 792
validated on BC with a predominance of gram-negative cocci in cluster showing 95.6% -100% of 793
sensitivity and 95.3% - 98.4% specificity for the detection of MRSA (97; 130; 140; 222). Discrepant 794
results are mainly explained by mixed culture or inhibition of the PCR. Negative MRSA detection is also 795
due to MREJ variants that contain the staphylococcal cassette chromosome mec (SCCmec) without the 796
mecA gene (219). 797
The StaphPlex system (Genaco Biomedical Products, Huntsville, AL, USA) is based on multiple PCRs 798
(18 target genes) and microarray analysis for the identification of staphylococci at species-level, for the 799
detection of resistance genes and for the detection of the Panton-Valentine leukocidin (PVL) with a time 800
to result of about 5 hours. The overall accuracy for the detection of staphylococci at the species level from 801
mono-bacterial or poly-bacterial culture was 91.7% when compared to conventional methods (230). The 802
StaphPlex exhibited 100% of sensitivity and 95.5% to 100.0% of specificity for the detection of the mecA 803
gene. Similar results were obtained for the detection of the PVL gene (230). 804
36
The FilmArray (Idaho Technology, Salt Lake City, UT, USA) is another PCR-based system that can 805
detect the S. aureus mecA gene as well as the vanA and vanB genes implicated in vancomycin resistance 806
in Enterococcus species (16; 192). The FilmArray system exhibits a high sensitivity and specificity 807
(100%) for the detection of the mecA gene from S. aureus positive culture with (16; 19; 196). Another 808
study reported 96% of sensitivity and 98.9% of specificity for the detection of the mecA gene (6). For the 809
detection of the vanA and vanB genes in Enterococcus spp., Blasche et al. reported only 85% of 810
specificity of the FilmArray system when compared to BC (19). In this study, the FilmArray detected a 811
van gene in 2 Enterococcus strains that where found vancomycin sensitive by culture based methods. The 812
second strain was actually an Enterococcus casseliflavus strain bearing the vanC gene. During the 813
development of the system, more resistance markers have been successfully tested (19). 814
The microarray of the Verigene gram-negative system (Northbrook, IL, USA) includes probes for the 815
detection of the mecA gene with 98.6% of sensitivity and 94.3% of specificity (26). This system also 816
contains probes for the detection of VRE with 100% of sensitivity and specificity in a prospective study 817
from positive BC (26). 818
The major weakness of the nucleic acid methods is to assign the resistance gene to the correct 819
microorganism in the case of mixed cultures. This is particularly limiting for the detection of the mecA 820
gene in a mix culture of S. aureus that would contain contaminating coagulase negative staphylococci for 821
instance (26). 822
Detection of resistance mechanisms of gram-negative bacteria 823
Gram-negative bacteria can exhibit resistance mechanisms to multiple antibiotics, representing a major 824
problem for the treatment of infections. In particular Enterobacteriaceae have developed multiple 825
resistance mechanisms to cephalosporins, such as ESBL. This had presented the carbapenem as 826
alternative for the treatments of severe infections involving Enterobacteriaceae and non-fermentative 827
bacteria such as Pseudomonas aeruginosa and Acinetobacter baumannii. However, resistance to 828
37
carbapenem may also occur. It might be caused by a reduced permeability with loss of porins, by the 829
over-expression of efflux pumps or alternatively by the production of carbapenem degrading enzymes, 830
namely carbapenemases. The emergence of carbapenem resistance strains and their spread worldwide is 831
alarming both from a therapeutic point of view and from an epidemiological point of view. Nucleic acid 832
detection systems are focusing on the blaKPC gene that encode for K. pneumoniae carbapenemase. This is 833
the case of the FilmArray system that includes primers targeting the blaKPC gene (6; 17; 19; 196) and the 834
Verigene gram-negative system that contain probes for the detection of the blaKPC (17). More studies 835
involving KPC positive strains are needed to determine the sensitivity and specificity of these systems. 836
Moreover, future tools should aim at detecting a broader range of genes encoding SBL and/or 837
carbapenemases. 838
2. Resistance detection from positive blood culture using MALDI-TOF MS 839
The use of the MALDI-TOF MS for the detection of resistant strains was first proposed for the distinction 840
between MRSA and MSSA colonies (68). However, recent developments are mainly directed toward 841
gram-negative bacteria in which The MALDI-TOF MS may detect antibiotics degrading proteins such as 842
carbapenemases. Practically a bacterial inoculum is incubated with ertapenem. The MALDI-TOF MS 843
analysis performed on the supernatant is used to determine the shift of two specific peaks associated with 844
the degradation of the carbapenem. This method has been validated on bacterial colonies with 100% of 845
sensitivity and 100% of specific for the detection of carbapenemase producing strains (240). Carvalhaes 846
and colleagues have proposed a method to be used directly from bacterial pellet obtain from positive BC, 847
which is able to identify 72.4% (21/29) of the carbapenemase-producing isolates after 4 h of incubation 848
(33). MALDI-TOF MS represents an alternative for the detection of carbapenemase from positive BC that 849
has the advantage, over PCR, to provide a phenotypic result. 850
38
3. Resistance detection directly from whole blood 851
As said earlier, the detection of resistance genes directly from blood would considerably impact the 852
choice of the therapy. Four primer set of the PLEX-ID system are used to detect the resistance cassettes 853
mecA, vanA and vanB and blaKPC, (70; 247). It has to be noted that the PLEX-ID system can also detect 854
the presence of the S. aureus Panton-Valentine leukocidin toxin (PVL) (247). PCR/ESI-MS was also 855
shown to be successful at identifying the mutations in gyrA and parC genes involved in Acinetobacter 856
baumannii resistance to quinolones (114; 115). In the case of mixed population, quantitative analysis 857
could help at associating a resistance gene to the corresponding pathogen. The MagicPlex system can 858
detect three resistance markers (mecA, vanA, vanB) directly from blood (152). From their study, Carrara 859
et al. failed to detect six MRSA strains out of ten using the MagicPlex system (30). The VYOO system 860
can detect five resistance genes: mecA, vanA, vanB β-lactamase blaSHV, β-lactamase blaCTX-M. However 861
there is a lack of studies reporting the sensitivity of this method. Finally, the SeptiFast system contains 862
primers for the detection of mecA only (table 1). 863
Although molecular methods have the ability to detect some resistance genes, more studies are required to 864
determine their performances and their clinical impact. In addition, increasing the number of available 865
resistance markers would be an advantage. 866
867
CONCLUSIONS AND FUTURE DIRECTIONS 868
Nucleic acid methods have succeeded to overcome most of the obstacle that limited the sensitivity of the 869
detection of pathogen from blood containing sample. The sensitivity and specificity are now adequate for 870
the use in diagnostic laboratory. 871
39
One major problem for the determination of the performances of new molecular methods is the fact that 872
BC associated with traditional identification methods remains the gold-standard. In this context, many 873
assays display extremely high analytic sensitivity but limited apparent specificity due to a limited 874
correlation with BC. This could be overcome by the use of other evaluation methods to challenge BC. In 875
their study, Bacconi and colleagues showed that 16S sequencing is not an appropriate method to be used 876
as reference for the evaluation of PCR/ESI-MS. Indeed, only 2 of the 35 samples positive using the 877
PCR/ESI-MS were found positive by 16S sequencing when the sequencing was performed on the same 878
material. Thus one of the limits to develop new methods for the diagnostic of BSI is that blood culture 879
remains the gold standard (13). To compare the performance of blood culture and other methods, clinical 880
presentations and local epidemiology should be taken into account in order to properly investigate 881
discrepant results (13; 137). Nevertheless the poor correlation between conventional BC-based methods 882
and some new molecular methods suggest that different pathogens are detected by these two methods and 883
that these diagnostic tools should be complementary. Moreover BC could not be replaced since the 884
availability of a strain in pure culture is mandatory to precisely test antibiotic susceptibility of bacteria. 885
Although new molecular methods are really appealing, further studies are still needed to determine their 886
impact on the management of patients with BSI (175). For instance, there is a lack of study that addresses 887
the impact on antibiotic stewardship. While most of the studies try to give with precision the 888
performances of the new molecular methods, it is not yet clear how these methods can be integrated in a 889
molecular laboratory. This is particularly true as none of the method would replace BC. Some practical 890
details have to be investigated. Thus, the number of samples, the number of venipuncture and the 891
frequency of the sampling for molecular diagnostic has to be addressed. Otherwise, informations 892
(routinely provided by the BC) on the likelihood of an infection (such as a possible catheter infection) 893
versus a contamination would be missing. 894
Molecular methods can also be helpful for organisms that are phenotypically closely related or for rare 895
organisms. However, because some genetically closely related organisms cannot be discriminated on the 896
40
basis of molecular methods such as E. coli/Shigella (59; 210) or group mitis streptococci/Streptococcus 897
pneumoniae (126; 224; 227), phenotypic distinction methods remain crucial. In the same manner the 898
accuracy of many molecular methods is based on their databases, which have to be maintained by adding 899
nucleic acid sequences or mass-spectrum obtained for new clinical isolates or rare organisms. For nucleic 900
acid methods, it is also crucial to achieve a follow-up of the emergence of SNP or mutations 901
(insertion/deletion) that could affect the hybridization of specific primers and probes. One solution would 902
be to rely on multiple targets (sequence or gene) for the identification of a given organism (120; 222). 903
As already stressed earlier the Gram staining remains mandatory for BC based methods since (i) it can 904
help at disclosing poly-microbial infection, (ii) it serves as a quality control and (iii) it sometimes helps 905
identification thanks to specific phenotypic traits. Nevertheless the presence of a single morphotype on 906
the Gram staining, does not exclude the presence of multiple organisms, since Enterobacteriaceae often 907
exhibit a similar Gram staining morphotype. 908
Regarding poly-microbial identification, the quantification provided by some nucleic-acids based 909
methods, can be helpful at determining the significance of the respective organisms. Another research 910
focus is the attempt to provide clinical scores that would predict the severity of bacteremia (78; 195). The 911
time to positivity of blood BC has been proposed as a criterion for BSI severity (7; 113; 129). Similarly, 912
quantification could provide some insight on the severity of the BSI. Thus, analysis of 250 whole-blood 913
samples from 20 adult patients (13 survivors and 7 nonsurvivors) with culture-proven MRSA showed that 914
the levels of mecA DNA was higher in the nonsurvivors (5.48 copies/ml) than in the survivors (4.58 log 915
copies/ml P= 0.003, two-tailed Mann-Whitney U test). This suggested that the level of mecA DNA in 916
blood could potentially be used to monitor MRSA bacteremia and evaluate responses to therapy (109). 917
Similarly a positive correlation has been found between the level of Streptococcus pneumoniae DNA 918
monitored by real-time PCR and the need of mechanical ventilation, the risk of septic shock and of death 919
(32; 198). Another study performed with the SeptiFast on 94 patients reported that the median cycle 920
threshold (Ct) value was 16.9 for patients with severe septic choc and 20.9 for patients with non-severe 921
41
sepsis and that Ct value <17.5 correlated with more positive blood culture and longer hospital stays (80; 922
266). 923
The quantification can also help at identifying the presence of a contaminant. In this context, the threshold 924
of the quantitative results becomes an important parameter to limit the number of false positive and false 925
negative results. Molecular solutions devoid of quantitative or at least semi-quantitative analysis will 926
render the interpretation of the positive results difficult. The interpretation of poly-microbial detection as 927
well as mono-bacterial ones is facilitated by the fact that PCR/ESI-MS gives is a quantitative analysis 928
performed using an internal control. This can also helps at identifying true positive versus contaminating 929
organisms (137) and at determining the relevance of each pathogen for mixed infections (75). Further 930
studies are required to determine if the quantification could also help at defining the severity of the 931
infection. 932
Regarding persisting bacteremia in the context of an ongoing treatment, molecular methods have been 933
able to rapidly detect new organisms that BC could not detect. However, nucleic-acid methods are not 934
adequate for the follow-up of persistent infections because they can detect the persistence of DNA from 935
dead organisms rather than true persisting organism. Therefore, many methods are now being developed 936
to detect only DNA from living organisms (1; 172). This could be of particular interest to monitor the 937
efficiency of an antibiotic treatment (29; 248). 938
Indication on the presence of resistance genes is an important added value of some molecular methods 939
being especially helpful at identifying risk of treatment failure. Development should focus in the 940
integration of more molecular markers. Molecular diagnosis may also detect the presence of virulence 941
genes, which helps predicting the severity of infection. Thus the StaphPlex system helps identifying 942
staphylococci at the species-level and detect resistance genes as well as the Panton-Valentine leukocidin 943
(PVL), excellent sensitivity (100%) and specificity (>95.5%) (230). 944
42
In conclusion, molecular diagnosis has significantly improved the diagnosis of BSI due to the reduction 945
of the time to result and to the high sensitivity and specificity. In particular the MALDI-TOF MS is a 946
revolution for the diagnosis of BSI from positive BC. To impact the management of patients suffering of 947
BSIs, microbiologist and clinicians should imagine new laboratories and algorithms that associate these 948
new culture-independent and culture dependant molecular methods with conventional methods in the aim 949
to get the benefits from both diagnosis methods. 950
951
952
43
REFERENCES 953
1. Aellen S, Que YA, Guignard B, Haenni M, Moreillon P. 2006. Detection of live and antibiotic-954
killed bacteria by quantitative real-time PCR of specific fragments of rRNA. Antimicrobial 955
agents and chemotherapy 50:1913-20 956
2. Akane A, Matsubara K, Nakamura H, Takahashi S, Kimura K. 1994. Identification of the heme 957
compound copurified with deoxyribonucleic acid (DNA) from bloodstains, a major inhibitor of 958
polymerase chain reaction (PCR) amplification. Journal of forensic sciences 39:362-72 959
3. Al-Soud WA, Jonsson LJ, Radstrom P. 2000. Identification and characterization of 960
immunoglobulin G in blood as a major inhibitor of diagnostic PCR. Journal of clinical 961
microbiology 38:345-50 962
4. Al-Soud WA, Radstrom P. 2001. Purification and characterization of PCR-inhibitory components 963
in blood cells. Journal of clinical microbiology 39:485-93 964
5. Alby K, Daniels LM, Weber DJ, Miller MB. 2013. Development of a treatment algorithm for 965
streptococci and enterococci from positive blood cultures identified with the Verigene Gram-966
positive blood culture assay. Journal of clinical microbiology 51:3869-71 967
6. Altun O, Almuhayawi M, Ullberg M, Ozenci V. 2013. Clinical evaluation of the FilmArray blood 968
culture identification panel in identification of bacteria and yeasts from positive blood culture 969
bottles. Journal of clinical microbiology 51:4130-6 970
7. Alvarez R, Vinas-Castillo L, Lepe-Jimenez JA, Garcia-Cabrera E, Cisneros-Herreros JM. 2012. 971
Time to positivity of blood culture association with clinical presentation, prognosis and ESBL-972
production in Escherichia coli bacteremia. European journal of clinical microbiology & 973
infectious diseases 31:2191-5 974
44
8. Anhalt JP, Fenselau C. 1975. Identification of bacteria using mass spectrometry. Anal. Chem 975
47:219–25 976
9. Arbique JC, Poyart C, Trieu-Cuot P, Quesne G, Carvalho Mda G, et al. 2004. Accuracy of 977
phenotypic and genotypic testing for identification of Streptococcus pneumoniae and description 978
of Streptococcus pseudopneumoniae sp. nov. Journal of clinical microbiology 42:4686-96 979
10. Arnold RJ, Karty JA, Ellington AD, Reilly JP. 1999. Monitoring the growth of a bacteria culture 980
by MALDI-MS of whole cells. Analytical chemistry 71:1990-6 981
11. Azevedo NF, Jardim T, Almeida C, Cerqueira L, Almeida AJ, et al. 2011. Application of flow 982
cytometry for the identification of Staphylococcus epidermidis by peptide nucleic acid 983
fluorescence in situ hybridization (PNA FISH) in blood samples. Antonie van Leeuwenhoek 984
100:463-70 985
12. Bacchetti De Gregoris T, Aldred N, Clare AS, Burgess JG. 2011. Improvement of phylum- and 986
class-specific primers for real-time PCR quantification of bacterial taxa. Journal of 987
microbiological methods 86:351-6 988
13. Bacconi A, Richmond GS, Baroldi MA, Laffler TG, Blyn LB, et al. 2014. Improved Sensitivity 989
for Molecular Detection of Bacteria and Candida in Blood. Journal of clinical microbiology 990
14. Beal SG, Ciurca J, Smith G, John J, Lee F, et al. 2013. Evaluation of the nanosphere verigene 991
gram-positive blood culture assay with the VersaTREK blood culture system and assessment of 992
possible impact on selected patients. Journal of clinical microbiology 51:3988-92 993
15. Bhatia NS, Farrell JJ, Sampath R, Ranken R, Rounds MA, et al. 2012. Identification of 994
Streptococcus intermedius central nervous system infection by use of PCR and electrospray 995
ionization mass spectrometry. Journal of clinical microbiology 50:4160-2 996
45
16. Bhatti MM, Boonlayangoor S, Beavis KG, Tesic V. 2014. Evaluation of FilmArray and Verigene 997
Systems for Rapid Identification of Positive Blood Cultures. Journal of clinical microbiology 998
52:3433-6 999
17. Bhatti MM, Boonlayangoor S, Beavis KG, Tesic V. 2014. Evaluation of FilmArray and Verigene 1000
Systems in the Rapid Identification of Positive Blood Cultures. Journal of clinical microbiology 1001
18. Bizzini A, Greub G. 2010. Matrix-assisted laser desorption ionization time-of-flight mass 1002
spectrometry, a revolution in clinical microbial identification. Clinical microbiology and infection 1003
16:1614-9 1004
19. Blaschke AJ, Heyrend C, Byington CL, Fisher MA, Barker E, et al. 2012. Rapid identification of 1005
pathogens from positive blood cultures by multiplex polymerase chain reaction using the 1006
FilmArray system. Diagnostic microbiology and infectious disease 74:349-55 1007
20. Bloos F, Bayer O, Sachse S, Straube E, Reinhart K, Kortgen A. 2013. Attributable costs of 1008
patients with candidemia and potential implications of polymerase chain reaction-based pathogen 1009
detection on antifungal therapy in patients with sepsis. Journal of critical care 28:2-8 1010
21. Bloos F, Hinder F, Becker K, Sachse S, Mekontso Dessap A, et al. 2010. A multicenter trial to 1011
compare blood culture with polymerase chain reaction in severe human sepsis. Intensive care 1012
medicine 36:241-7 1013
22. Bloos F, Sachse S, Kortgen A, Pletz MW, Lehmann M, et al. 2012. Evaluation of a polymerase 1014
chain reaction assay for pathogen detection in septic patients under routine condition: an 1015
observational study. PloS one 7:e46003 1016
46
23. Blyn LB, Hall TA, Libby B, Ranken R, Sampath R, et al. 2008. Rapid detection and molecular 1017
serotyping of adenovirus by use of PCR followed by electrospray ionization mass spectrometry. 1018
Journal of clinical microbiology 46:644-51 1019
24. Braun SD, Monecke S, Thurmer A, Ruppelt A, Makarewicz O, et al. 2014. Rapid identification of 1020
carbapenemase genes in gram-negative bacteria with an oligonucleotide microarray-based assay. 1021
PloS one 9:e102232 1022
25. Bravo D, Blanquer J, Tormo M, Aguilar G, Borras R, et al. 2011. Diagnostic accuracy and 1023
potential clinical value of the LightCycler SeptiFast assay in the management of bloodstream 1024
infections occurring in neutropenic and critically ill patients. International journal of infectious 1025
diseases : IJID 15:e326-31 1026
26. Buchan BW, Ginocchio CC, Manii R, Cavagnolo R, Pancholi P, et al. 2013. Multiplex 1027
identification of gram-positive bacteria and resistance determinants directly from positive blood 1028
culture broths: evaluation of an automated microarray-based nucleic acid test. PLoS medicine 1029
10:e1001478 1030
27. Byrnes JJ, Downey KM, Esserman L, So AG. 1975. Mechanism of hemin inhibition of erythroid 1031
cytoplasmic DNA polymerase. Biochemistry 14:796-9 1032
28. Cain SE, Kohn J, Bookstaver PB, Albrecht H, Al-Hasan MN. 2015. Stratification of the impact of 1033
inappropriate empirical antimicrobial therapy for gram-negative bloodstream infections by 1034
predicted prognosis. Antimicrobial agents and chemotherapy 59:245-50 1035
29. Cangelosi GA, Meschke JS. 2014. Dead or Alive: Molecular Assessment of Microbial Viability. 1036
Applied and environmental microbiology 1037
47
30. Carrara L, Navarro F, Turbau M, Seres M, Moran I, et al. 2013. Molecular diagnosis of 1038
bloodstream infections with a new dual-priming oligonucleotide-based multiplex PCR assay. 1039
Journal of medical microbiology 62:1673-9 1040
31. Carretto E, Bardaro M, Russello G, Mirra M, Zuelli C, Barbarini D. 2013. Comparison of the 1041
Staphylococcus QuickFISH BC test with the tube coagulase test performed on positive blood 1042
cultures for evaluation and application in a clinical routine setting. Journal of clinical 1043
microbiology 51:131-5 1044
32. Carrol ED, Guiver M, Nkhoma S, Mankhambo LA, Marsh J, et al. 2007. High pneumococcal 1045
DNA loads are associated with mortality in Malawian children with invasive pneumococcal 1046
disease. The Pediatric infectious disease journal 26:416-22 1047
33. Carvalhaes CG, Cayo R, Visconde MF, Barone T, Frigatto EA, et al. 2014. Detection of 1048
carbapenemase activity directly from blood culture vials using MALDI-TOF MS: a quick answer 1049
for the right decision. The Journal of antimicrobial chemotherapy 69:2132-6 1050
34. Casalta JP, Gouriet F, Roux V, Thuny F, Habib G, Raoult D. 2009. Evaluation of the LightCycler 1051
SeptiFast test in the rapid etiologic diagnostic of infectious endocarditis. European journal of 1052
clinical microbiology & infectious diseases 28:569-73 1053
35. Chang SS, Hsieh WH, Liu TS, Lee SH, Wang CH, et al. 2013. Multiplex PCR system for rapid 1054
detection of pathogens in patients with presumed sepsis - a systemic review and meta-analysis. 1055
PloS one 8:e62323 1056
36. Chapin K, Musgnug M. 2003. Evaluation of three rapid methods for the direct identification of 1057
Staphylococcus aureus from positive blood cultures. Journal of clinical microbiology 41:4324-7 1058
48
37. Chaubey VP, Pitout JD, Dalton B, Ross T, Church DL, et al. 2010. Clinical outcome of empiric 1059
antimicrobial therapy of bacteremia due to extended-spectrum beta-lactamase producing 1060
Escherichia coli and Klebsiella pneumoniae. BMC research notes 3:116 1061
38. Chen J, Fu Y, Ju L, Miao X, Shen Y, et al. 2014. Detection and identification of viral pathogens 1062
in patients with hand, foot, and mouth disease by multilocus PCR, reverse-transcription PCR and 1063
electrospray ionization mass spectrometry. Journal of clinical virology59:115-9 1064
39. Chen KF, Rothman RE, Ramachandran P, Blyn L, Sampath R, et al. 2011. Rapid identification 1065
viruses from nasal pharyngeal aspirates in acute viral respiratory infections by RT-PCR and 1066
electrospray ionization mass spectrometry. Journal of virological methods 173:60-6 1067
40. Cherkaoui A, Ceroni D, Emonet S, Lefevre Y, Schrenzel J. 2009. Molecular diagnosis of 1068
Kingella kingae osteoarticular infections by specific real-time PCR assay. Journal of medical 1069
microbiology 58:65-8 1070
41. Cherkaoui A, Emonet S, Ceroni D, Candolfi B, Hibbs J, et al. 2009. Development and validation 1071
of a modified broad-range 16S rDNA PCR for diagnostic purposes in clinical microbiology. 1072
Journal of microbiological methods 79:227-31 1073
42. Christner M, Rohde H, Wolters M, Sobottka I, Wegscheider K, Aepfelbacher M. 2010. Rapid 1074
identification of bacteria from positive blood culture bottles by use of matrix-assisted laser 1075
desorption-ionization time of flight mass spectrometry fingerprinting. Journal of clinical 1076
microbiology 48:1584-91 1077
43. Clarridge JE, 3rd. 2004. Impact of 16S rRNA gene sequence analysis for identification of bacteria 1078
on clinical microbiology and infectious diseases. Clinical microbiology reviews 17:840-62, table 1079
of contents 1080
49
44. Claydon MA, Davey SN, Edwards-Jones V, Gordon DB. 1996. The rapid identification of intact 1081
microorganisms using mass spectrometry. Nature biotechnology 14:1584-6 1082
45. Clerc O, Prod'hom G, Senn L, Jaton K, Zanetti G, et al. 2014. Matrix-assisted laser desorption 1083
ionization time-of-flight mass spectrometry and PCR-based rapid diagnosis of Staphylococcus 1084
aureus bacteraemia. Clinical microbiology and infection 20:355-60 1085
46. Clerc O, Prod'hom G, Vogne C, Bizzini A, Calandra T, Greub G. 2013. Impact of matrix-assisted 1086
laser desorption ionization time-of-flight mass spectrometry on the clinical management of 1087
patients with Gram-negative bacteremia: a prospective observational study. Clinical infectious 1088
diseases 56:1101-7 1089
47. Cleven BE, Palka-Santini M, Gielen J, Meembor S, Kronke M, Krut O. 2006. Identification and 1090
characterization of bacterial pathogens causing bloodstream infections by DNA microarray. 1091
Journal of clinical microbiology 44:2389-97 1092
48. Cordey S, Thomas Y, Suter P, Kaiser L. 2012. Pilot Evaluation of RT-PCR/Electrospray 1093
Ionization Mass Spectrometry (PLEX-ID/Flu assay) on Influenza-Positive Specimens. The open 1094
virology journal 6:64-7 1095
49. Cosgrove SE, Qi Y, Kaye KS, Harbarth S, Karchmer AW, Carmeli Y. 2005. The impact of 1096
methicillin resistance in Staphylococcus aureus bacteremia on patient outcomes: mortality, length 1097
of stay, and hospital charges. Infection control and hospital epidemiology : the official journal of 1098
the Society of Hospital Epidemiologists of America 26:166-74 1099
50. Cosgrove SE, Sakoulas G, Perencevich EN, Schwaber MJ, Karchmer AW, Carmeli Y. 2003. 1100
Comparison of mortality associated with methicillin-resistant and methicillin-susceptible 1101
Staphylococcus aureus bacteremia: a meta-analysis. Clinical infectious diseases 36:53-9 1102
50
51. Croxatto A, Prod'hom G, Durussel C, Greub G. 2014. Preparation of a blood culture pellet for 1103
rapid bacterial identification and antibiotic susceptibility testing. Journal of visualized 1104
experiments : JoVE 1105
52. Croxatto A, Prod'hom G, Greub G. 2012. Applications of MALDI-TOF mass spectrometry in 1106
clinical diagnostic microbiology. FEMS microbiology reviews 36:380-407 1107
53. Dark P, Wilson C, Blackwood B, McAuley DF, Perkins GD, et al. 2012. Accuracy of 1108
LightCycler(R) SeptiFast for the detection and identification of pathogens in the blood of patients 1109
with suspected sepsis: a systematic review protocol. BMJ open 2:e000392 1110
54. Davies J, Gordon CL, Tong SY, Baird RW, Davis JS. 2012. Impact of results of a rapid 1111
Staphylococcus aureus diagnostic test on prescribing of antibiotics for patients with clustered 1112
gram-positive cocci in blood cultures. Journal of clinical microbiology 50:2056-8 1113
55. Deck MK, Anderson ES, Buckner RJ, Colasante G, Coull JM, et al. 2012. Multicenter evaluation 1114
of the Staphylococcus QuickFISH method for simultaneous identification of Staphylococcus 1115
aureus and coagulase-negative staphylococci directly from blood culture bottles in less than 30 1116
minutes. Journal of clinical microbiology 50:1994-8 1117
56. Deck MK, Anderson ES, Buckner RJ, Colasante G, Davis TE, et al. 2014. Rapid detection of 1118
Enterococcus spp. direct from blood culture bottles using Enterococcus QuickFISH method: a 1119
multicenter investigation. Diagnostic microbiology and infectious disease 78:338-42 1120
57. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, et al. 2008. Surviving Sepsis Campaign: 1121
international guidelines for management of severe sepsis and septic shock: 2008. Intensive care 1122
medicine 34:17-60 1123
51
58. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, et al. 2013. Surviving sepsis 1124
campaign: international guidelines for management of severe sepsis and septic shock: 2012. 1125
Critical care medicine 41:580-637 1126
59. Deng J, Fu L, Wang R, Yu N, Ding X, et al. 2014. Comparison of MALDI-TOF MS, gene 1127
sequencing and the Vitek 2 for identification of seventy-three clinical isolates of enteropathogens. 1128
Journal of thoracic disease 6:539-44 1129
60. Dodemont M, De Mendonca R, Nonhoff C, Roisin S, Denis O. 2014. Performance of the 1130
verigene gram-negative blood culture assay for rapid detection of bacteria and resistance 1131
determinants. Journal of clinical microbiology 52:3085-7 1132
61. Drancourt M. 2010. Detection of microorganisms in blood specimens using matrix-assisted laser 1133
desorption ionization time-of-flight mass spectrometry: a review. Clinical microbiology and 1134
infection 16:1620-5 1135
62. Duncan DD, Vogler AJ, Wolcott MJ, Li F, Sarovich DS, et al. 2013. Identification and typing of 1136
Francisella tularensis with a highly automated genotyping assay. Letters in applied microbiology 1137
56:128-34 1138
63. Ecker DJ, Massire C, Blyn LB, Hofstadler SA, Hannis JC, et al. 2009. Molecular genotyping of 1139
microbes by multilocus PCR and mass spectrometry: a new tool for hospital infection control and 1140
public health surveillance. Methods in molecular biology 551:71-87 1141
64. Ecker DJ, Sampath R, Blyn LB, Eshoo MW, Ivy C, et al. 2005. Rapid identification and strain-1142
typing of respiratory pathogens for epidemic surveillance. Proceedings of the National Academy 1143
of Sciences of the United States of America 102:8012-7 1144
52
65. Ecker DJ, Sampath R, Li H, Massire C, Matthews HE, et al. 2010. New technology for rapid 1145
molecular diagnosis of bloodstream infections. Expert review of molecular diagnostics 10:399-1146
415 1147
66. Ecker DJ, Sampath R, Massire C, Blyn LB, Hall TA, et al. 2008. Ibis T5000: a universal 1148
biosensor approach for microbiology. Nature reviews. Microbiology 6:553-8 1149
67. Ecker JA, Massire C, Hall TA, Ranken R, Pennella TT, et al. 2006. Identification of 1150
Acinetobacter species and genotyping of Acinetobacter baumannii by multilocus PCR and mass 1151
spectrometry. Journal of clinical microbiology 44:2921-32 1152
68. Edwards-Jones V, Claydon MA, Evason DJ, Walker J, Fox AJ, Gordon DB. 2000. Rapid 1153
discrimination between methicillin-sensitive and methicillin-resistant Staphylococcus aureus by 1154
intact cell mass spectrometry. Journal of medical microbiology 49:295-300 1155
69. Elenitoba-Johnson KS, Bohling SD, Wittwer CT, King TC. 2001. Multiplex PCR by multicolor 1156
fluorimetry and fluorescence melting curve analysis. Nature medicine 7:249-53 1157
70. Endimiani A, Hujer KM, Hujer AM, Sampath R, Ecker DJ, Bonomo RA. 2010. Rapid 1158
identification of bla KPC-possessing Enterobacteriaceae by PCR/electrospray ionization-mass 1159
spectrometry. The Journal of antimicrobial chemotherapy 65:1833-4 1160
71. Engemann JJ, Carmeli Y, Cosgrove SE, Fowler VG, Bronstein MZ, et al. 2003. Adverse clinical 1161
and economic outcomes attributable to methicillin resistance among patients with Staphylococcus 1162
aureus surgical site infection. Clinical infectious diseases 36:592-8 1163
72. Erali M, Pounder JI, Woods GL, Petti CA, Wittwer CT. 2006. Multiplex single-color PCR with 1164
amplicon melting analysis for identification of Aspergillus species. Clinical chemistry 52:1443-5 1165
53
73. Eshoo MW, Crowder CD, Li H, Matthews HE, Meng S, et al. 2010. Detection and identification 1166
of Ehrlichia species in blood by use of PCR and electrospray ionization mass spectrometry. 1167
Journal of clinical microbiology 48:472-8 1168
74. Farina C, Perin S, Andreoni S, Conte M, Fazii P, et al. 2012. Evaluation of the peptide nucleic 1169
acid fluorescence in situ hybridisation technology for yeast identification directly from positive 1170
blood cultures: an Italian experience. Mycoses 55:388-92 1171
75. Farrell JJ, Sampath R, Ecker DJ, Bonomo RA. 2013. "Salvage microbiology": detection of 1172
bacteria directly from clinical specimens following initiation of antimicrobial treatment. PloS one 1173
8:e66349 1174
76. Farrell JJ, Tsung AJ, Flier L, Martinez DL, Beam SB, et al. 2013. PCR and electrospray 1175
ionization mass spectrometry for detection of persistent enterococcus faecalis in cerebrospinal 1176
fluid following treatment of postoperative ventriculitis. Journal of clinical microbiology 51:3464-1177
6 1178
77. Fatkenheuer G, Preuss M, Salzberger B, Schmeisser N, Cornely OA, et al. 2004. Long-term 1179
outcome and quality of care of patients with Staphylococcus aureus bacteremia. European 1180
journal of clinical microbiology & infectious diseases 23:157-62 1181
78. Feldman C, Alanee S, Yu VL, Richards GA, Ortqvist A, et al. 2009. Severity of illness scoring 1182
systems in patients with bacteraemic pneumococcal pneumonia: implications for the intensive 1183
care unit care. Clinical microbiology and infection 15:850-7 1184
79. Fenollar F, Raoult D. 2007. Molecular diagnosis of bloodstream infections caused by non-1185
cultivable bacteria. International journal of antimicrobial agents 30 Suppl 1:S7-15 1186
54
80. Fernandez-Cruz A, Marin M, Kestler M, Alcala L, Rodriguez-Creixems M, Bouza E. 2013. The 1187
value of combining blood culture and SeptiFast data for predicting complicated bloodstream 1188
infections caused by Gram-positive bacteria or Candida species. Journal of clinical microbiology 1189
51:1130-6 1190
81. Fernandez AL, Varela E, Martinez L, Martinez A, Sierra J, et al. 2010. Evaluation of a multiplex 1191
real-time PCR assay for detecting pathogens in cardiac valve tissue in patients with endocarditis. 1192
Revista espanola de cardiologia 63:1205-8 1193
82. Ferrer R, Martin-Loeches I, Phillips G, Osborn TM, Townsend S, et al. 2014. Empiric antibiotic 1194
treatment reduces mortality in severe sepsis and septic shock from the first hour: results from a 1195
guideline-based performance improvement program*. Critical care medicine 42:1749-55 1196
83. Ferroni A, Suarez S, Beretti JL, Dauphin B, Bille E, et al. 2010. Real-time identification of 1197
bacteria and Candida species in positive blood culture broths by matrix-assisted laser desorption 1198
ionization-time of flight mass spectrometry. Journal of clinical microbiology 48:1542-8 1199
84. Fitting C, Parlato M, Adib-Conquy M, Memain N, Philippart F, et al. 2012. DNAemia detection 1200
by multiplex PCR and biomarkers for infection in systemic inflammatory response syndrome 1201
patients. PloS one 7:e38916 1202
85. Forrest GN. 2007. PNA FISH: present and future impact on patient management. Expert review 1203
of molecular diagnostics 7:231-6 1204
86. Forrest GN, Mehta S, Weekes E, Lincalis DP, Johnson JK, Venezia RA. 2006. Impact of rapid in 1205
situ hybridization testing on coagulase-negative staphylococci positive blood cultures. The 1206
Journal of antimicrobial chemotherapy 58:154-8 1207
55
87. Forrest GN, Roghmann MC, Toombs LS, Johnson JK, Weekes E, et al. 2008. Peptide nucleic 1208
acid fluorescent in situ hybridization for hospital-acquired enterococcal bacteremia: delivering 1209
earlier effective antimicrobial therapy. Antimicrobial agents and chemotherapy 52:3558-63 1210
88. Fuglsang-Damgaard D, Nielsen CH, Mandrup E, Fuursted K. 2011. The use of Gram stain and 1211
matrix-assisted laser desorption ionization time-of-flight mass spectrometry on positive blood 1212
culture: synergy between new and old technology. APMIS : acta pathologica, microbiologica, et 1213
immunologica Scandinavica 119:681-8 1214
89. Gaibani P, Mariconti M, Bua G, Bonora S, Sassera D, et al. 2013. Development of a broad-range 1215
23S rDNA real-time PCR assay for the detection and quantification of pathogenic bacteria in 1216
human whole blood and plasma specimens. BioMed research international 2013:264651 1217
90. Garey KW, Rege M, Pai MP, Mingo DE, Suda KJ, et al. 2006. Time to initiation of fluconazole 1218
therapy impacts mortality in patients with candidemia: a multi-institutional study. Clinical 1219
infectious diseases 43:25-31 1220
91. Gariani K, Rougemont M, Renzi G, Hibbs J, Emonet S, Schrenzel J. 2014. Fulminant atypical 1221
Cryptococcus neoformans pneumonia confirmed by PLEX-ID. International journal of infectious 1222
diseases : IJID 22:17-8 1223
92. Goldberg E, Paul M, Talker O, Samra Z, Raskin M, et al. 2010. Co-trimoxazole versus 1224
vancomycin for the treatment of methicillin-resistant Staphylococcus aureus bacteraemia: a 1225
retrospective cohort study. The Journal of antimicrobial chemotherapy 65:1779-83 1226
93. Gonzalez V, Padilla E, Gimenez M, Vilaplana C, Perez A, et al. 2004. Rapid diagnosis of 1227
Staphylococcus aureus bacteremia using S. aureus PNA FISH. European journal of clinical 1228
microbiology & infectious diseases 23:396-8 1229
56
94. Greub G, Lepidi H, Rovery C, Casalta JP, Habib G, et al. 2005. Diagnosis of infectious 1230
endocarditis in patients undergoing valve surgery. The American journal of medicine 118:230-8 1231
95. Grif K, Fille M, Wurzner R, Weiss G, Lorenz I, et al. 2012. Rapid detection of bloodstream 1232
pathogens by real-time PCR in patients with sepsis. Wiener klinische Wochenschrift 124:266-70 1233
96. Grif K, Heller I, Prodinger WM, Lechleitner K, Lass-Florl C, Orth D. 2012. Improvement of 1234
detection of bacterial pathogens in normally sterile body sites with a focus on orthopedic samples 1235
by use of a commercial 16S rRNA broad-range PCR and sequence analysis. Journal of clinical 1236
microbiology 50:2250-4 1237
97. Grobner S, Dion M, Plante M, Kempf VA. 2009. Evaluation of the BD GeneOhm StaphSR assay 1238
for detection of methicillin-resistant and methicillin-susceptible Staphylococcus aureus isolates 1239
from spiked positive blood culture bottles. Journal of clinical microbiology 47:1689-94 1240
98. Guido M, Quattrocchi M, Zizza A, Pasanisi G, Pavone V, et al. 2012. Molecular approaches in 1241
the diagnosis of sepsis in neutropenic patients with haematological malignances. Journal of 1242
preventive medicine and hygiene 53:104-8 1243
99. Haag H, Locher F, Nolte O. 2013. Molecular diagnosis of microbial aetiologies using SepsiTest 1244
in the daily routine of a diagnostic laboratory. Diagnostic microbiology and infectious disease 1245
76:413-8 1246
100. Hall L, Le Febre KM, Deml SM, Wohlfiel SL, Wengenack NL. 2012. Evaluation of the Yeast 1247
Traffic Light PNA FISH probes for identification of Candida species from positive blood 1248
cultures. Journal of clinical microbiology 50:1446-8 1249
57
101. Hall TA, Sampath R, Blyn LB, Ranken R, Ivy C, et al. 2009. Rapid molecular genotyping and 1250
clonal complex assignment of Staphylococcus aureus isolates by PCR coupled to electrospray 1251
ionization-mass spectrometry. Journal of clinical microbiology 47:1733-41 1252
102. Hannis JC, Manalili SM, Hall TA, Ranken R, White N, et al. 2008. High-resolution genotyping 1253
of Campylobacter species by use of PCR and high-throughput mass spectrometry. Journal of 1254
clinical microbiology 46:1220-5 1255
103. Harris DM, Hata DJ. 2013. Rapid identification of bacteria and Candida using PNA-FISH from 1256
blood and peritoneal fluid cultures: a retrospective clinical study. Annals of clinical microbiology 1257
and antimicrobials 12:2 1258
104. Hartmann H, Stender H, Schafer A, Autenrieth IB, Kempf VA. 2005. Rapid identification of 1259
Staphylococcus aureus in blood cultures by a combination of fluorescence in situ hybridization 1260
using peptide nucleic acid probes and flow cytometry. Journal of clinical microbiology 43:4855-7 1261
105. Henry NK, McLimans CA, Wright AJ, Thompson RL, Wilson WR, Washington JA, 2nd. 1983. 1262
Microbiological and clinical evaluation of the isolator lysis-centrifugation blood culture tube. 1263
Journal of clinical microbiology 17:864-9 1264
106. Hensley DM, Tapia R, Encina Y. 2009. An evaluation of the advandx Staphylococcus 1265
aureus/CNS PNA FISH assay. Clinical laboratory science : journal of the American Society for 1266
Medical Technology 22:30-3 1267
107. Heo NY, Lim YS, An J, Ko SY, Oh HB. 2012. Multiplex polymerase chain reaction test for the 1268
diagnosis of acute viral hepatitis A. Clinical and molecular hepatology 18:397-403 1269
108. Hill JT, Tran KD, Barton KL, Labreche MJ, Sharp SE. 2014. Evaluation of the Nanosphere-1270
Verigene BC-GN for the direct identification of Gram negative bacilli and antibiotic resistance 1271
58
markers from positive blood cultures and the potential impact for more rapid antibiotic 1272
interventions. Journal of clinical microbiology 1273
109. Ho YC, Chang SC, Lin SR, Wang WK. 2009. High levels of mecA DNA detected by a 1274
quantitative real-time PCR assay are associated with mortality in patients with methicillin-1275
resistant Staphylococcus aureus bacteremia. Journal of clinical microbiology 47:1443-51 1276
110. Holland RD, Wilkes JG, Rafii F, Sutherland JB, Persons CC, et al. 1996. Rapid identification of 1277
intact whole bacteria based on spectral patterns using matrix-assisted laser desorption/ionization 1278
with time-of-flight mass spectrometry. Rapid communications in mass spectrometry : RCM 1279
10:1227-32 1280
111. Hoorfar J, Wolffs P, Radstrom P. 2004. Diagnostic PCR: validation and sample preparation are 1281
two sides of the same coin. APMIS : acta pathologica, microbiologica, et immunologica 1282
Scandinavica 112:808-14 1283
112. Horz HP, Scheer S, Vianna ME, Conrads G. 2010. New methods for selective isolation of 1284
bacterial DNA from human clinical specimens. Anaerobe 16:47-53 1285
113. Huang L, Zhang YY, Sun LY. 2013. Time to positivity of blood culture can predict different 1286
Candida species instead of pathogen concentration in candidemia. European journal of clinical 1287
microbiology & infectious diseases 32:917-22 1288
114. Hujer KM, Hujer AM, Endimiani A, Thomson JM, Adams MD, et al. 2009. Rapid determination 1289
of quinolone resistance in Acinetobacter spp. Journal of clinical microbiology 47:1436-42 1290
115. Hujer KM, Hujer AM, Hulten EA, Bajaksouzian S, Adams JM, et al. 2006. Analysis of antibiotic 1291
resistance genes in multidrug-resistant Acinetobacter sp. isolates from military and civilian 1292
59
patients treated at the Walter Reed Army Medical Center. Antimicrobial agents and 1293
chemotherapy 50:4114-23 1294
116. Idelevich EA, Schule I, Grunastel B, Wullenweber J, Peters G, Becker K. 2014. Rapid 1295
identification of microorganisms from positive blood cultures by MALDI-TOF mass 1296
spectrometry subsequent to very short-term incubation on solid medium. Clinical microbiology 1297
and infection 20:1001-6 1298
117. Ishikawa H, Kutsukake E, Chiba K, Fukui T, Matsumoto T. 2011. The performance of the BD 1299
geneOhm MRSA assay for MRSA isolated from clinical patients in Japan, including the effects 1300
of specimen contamination and ways to improve it. Journal of infection and chemotherapy : 1301
official journal of the Japan Society of Chemotherapy 17:214-8 1302
118. Jacob D, Sauer U, Housley R, Washington C, Sannes-Lowery K, et al. 2012. Rapid and high-1303
throughput detection of highly pathogenic bacteria by Ibis PLEX-ID technology. PloS one 1304
7:e39928 1305
119. Jacobs MR, Bajaksouzian S, Bonomo RA, Good CE, Windau AR, et al. 2009. Occurrence, 1306
distribution, and origins of Streptococcus pneumoniae Serotype 6C, a recently recognized 1307
serotype. Journal of clinical microbiology 47:64-72 1308
120. Jaton K, Ninet B, Bille J, Greub G. 2010. False-negative PCR result due to gene polymorphism: 1309
the example of Neisseria meningitidis. Journal of clinical microbiology 48:4590-1 1310
121. Jeng K, Gaydos CA, Blyn LB, Yang S, Won H, et al. 2012. Comparative analysis of two broad-1311
range PCR assays for pathogen detection in positive-blood-culture bottles: PCR-high-resolution 1312
melting analysis versus PCR-mass spectrometry. Journal of clinical microbiology 50:3287-92 1313
60
122. Jordana-Lluch E, Carolan HE, Gimenez M, Sampath R, Ecker DJ, et al. 2013. Rapid diagnosis of 1314
bloodstream infections with PCR followed by mass spectrometry. PloS one 8:e62108 1315
123. Jordana-Lluch E, Gimenez M, Quesada MD, Ausina V, Martro E. 2014. Improving the diagnosis 1316
of bloodstream infections: PCR coupled with mass spectrometry. BioMed research international 1317
2014:501214 1318
124. Josefson P, Stralin K, Ohlin A, Ennefors T, Dragsten B, et al. 2011. Evaluation of a commercial 1319
multiplex PCR test (SeptiFast) in the etiological diagnosis of community-onset bloodstream 1320
infections. European journal of clinical microbiology & infectious diseases 30:1127-34 1321
125. Kaleta EJ, Clark AE, Johnson DR, Gamage DC, Wysocki VH, et al. 2011. Use of PCR coupled 1322
with electrospray ionization mass spectrometry for rapid identification of bacterial and yeast 1323
bloodstream pathogens from blood culture bottles. Journal of clinical microbiology 49:345-53 1324
126. Karpanoja P, Harju I, Rantakokko-Jalava K, Haanpera M, Sarkkinen H. 2014. Evaluation of two 1325
matrix-assisted laser desorption ionization-time of flight mass spectrometry systems for 1326
identification of viridans group streptococci. European journal of clinical microbiology & 1327
infectious diseases 33:779-88 1328
127. Kasper DC, Altiok I, Mechtler TP, Bohm J, Straub J, et al. 2013. Molecular detection of late-1329
onset neonatal sepsis in premature infants using small blood volumes: proof-of-concept. 1330
Neonatology 103:268-73 1331
128. Khot PD, Fisher MA. 2013. Novel approach for differentiating Shigella species and Escherichia 1332
coli by matrix-assisted laser desorption ionization-time of flight mass spectrometry. Journal of 1333
clinical microbiology 51:3711-6 1334
61
129. Kim J, Gregson DB, Ross T, Laupland KB. 2010. Time to blood culture positivity in 1335
Staphylococcus aureus bacteremia: association with 30-day mortality. The Journal of infection 1336
61:197-204 1337
130. Kimura T, Komori T, Hirose Y, Kurahashi S, Yamada Y, et al. 2009. Evaluation of the MRSA 1338
rapid detection assay (BD GeneOhm MRSA detection kit) by a real-time PCR. Rinsho byori. The 1339
Japanese journal of clinical pathology 57:425-30 1340
131. Kramski M, Gaeguta AJ, Lichtfuss GF, Rajasuriar R, Crowe SM, et al. 2011. Novel sensitive 1341
real-time PCR for quantification of bacterial 16S rRNA genes in plasma of HIV-infected patients 1342
as a marker for microbial translocation. Journal of clinical microbiology 49:3691-3 1343
132. Kreger BE, Craven DE, Carling PC, McCabe WR. 1980. Gram-negative bacteremia. III. 1344
Reassessment of etiology, epidemiology and ecology in 612 patients. The American journal of 1345
medicine 68:332-43 1346
133. Kreger BE, Craven DE, McCabe WR. 1980. Gram-negative bacteremia. IV. Re-evaluation of 1347
clinical features and treatment in 612 patients. The American journal of medicine 68:344-55 1348
134. Kuhn C, Disque C, Muhl H, Orszag P, Stiesch M, Haverich A. 2011. Evaluation of commercial 1349
universal rRNA gene PCR plus sequencing tests for identification of bacteria and fungi associated 1350
with infectious endocarditis. Journal of clinical microbiology 49:2919-23 1351
135. L AM, Hardick J, Jeng K, Gaydos CA. 2013. Evaluation of the Pan Influenza detection kit 1352
utilizing the PLEX-ID and influenza samples from the 2011 respiratory season. Journal of 1353
virological methods 193:173-6 1354
136. La Scola B, Raoult D. 2009. Direct identification of bacteria in positive blood culture bottles by 1355
matrix-assisted laser desorption ionisation time-of-flight mass spectrometry. PloS one 4:e8041 1356
62
137. Laffler TG, Cummins LL, McClain CM, Quinn CD, Toro MA, et al. 2013. Enhanced diagnostic 1357
yields of bacteremia and candidemia in blood specimens by PCR-electrospray ionization mass 1358
spectrometry. Journal of clinical microbiology 51:3535-41 1359
138. Lai C, Lam W, Lee MP. 2014. A case of Mycobacterium sherrisii pneumonia diagnosed by 1360
PCR/ESI-MS method. International journal of infectious diseases : IJID 25:119-21 1361
139. Lamoth F, Jaton K, Prod'hom G, Senn L, Bille J, et al. 2010. Multiplex blood PCR in 1362
combination with blood cultures for improvement of microbiological documentation of infection 1363
in febrile neutropenia. Journal of clinical microbiology 48:3510-6 1364
140. Laudat P, Demondion E, Jouannet C, Charron J, Chillou C, et al. 2012. [Detection of 1365
Staphylococcus aureus resistant to methicillin (MRSA) by molecular biology (Cepheid 1366
GeneXpert IL, GeneOhm BD, Roche LightCycler, Hyplex Evigene I2A) versus screening by 1367
culture: Economic and practical strategy for the laboratory]. Pathologie-biologie 60:208-13 1368
141. Lefort A, Chartier L, Sendid B, Wolff M, Mainardi JL, et al. 2012. Diagnosis, management and 1369
outcome of Candida endocarditis. Clinical microbiology and infection 18:E99-E109 1370
142. Legoff J, Feghoul L, Mercier-Delarue S, Dalle JH, Scieux C, et al. 2013. Broad-range PCR-1371
electrospray ionization mass spectrometry for detection and typing of adenovirus and other 1372
opportunistic viruses in stem cell transplant patients. Journal of clinical microbiology 51:4186-92 1373
143. Lehmann LE, Hauser S, Malinka T, Klaschik S, Weber SU, et al. 2011. Rapid qualitative urinary 1374
tract infection pathogen identification by SeptiFast real-time PCR. PloS one 6:e17146 1375
144. Leibovici L, Shraga I, Drucker M, Konigsberger H, Samra Z, Pitlik SD. 1998. The benefit of 1376
appropriate empirical antibiotic treatment in patients with bloodstream infection. Journal of 1377
internal medicine 244:379-86 1378
63
145. Leitner E, Kessler HH, Spindelboeck W, Hoenigl M, Putz-Bankuti C, et al. 2013. Comparison of 1379
two molecular assays with conventional blood culture for diagnosis of sepsis. Journal of 1380
microbiological methods 92:253-5 1381
146. Leli C, Cardaccia A, D'Alo F, Ferri C, Bistoni F, Mencacci A. 2014. A prediction model for real-1382
time PCR results in blood samples from febrile patients with suspected sepsis. Journal of medical 1383
microbiology 63:649-58 1384
147. Leli C, Moretti A, Pasticci MB, Cenci E, Bistoni F, Mencacci A. 2014. A commercially available 1385
multiplex real-time PCR for detection of pathogens in cardiac valves from patients with infective 1386
endocarditis. Diagnostic microbiology and infectious disease 79:98-101 1387
148. Levy MM, Artigas A, Phillips GS, Rhodes A, Beale R, et al. 2012. Outcomes of the Surviving 1388
Sepsis Campaign in intensive care units in the USA and Europe: a prospective cohort study. The 1389
Lancet infectious diseases 12:919-24 1390
149. Levy MM, Dellinger RP, Townsend SR, Linde-Zwirble WT, Marshall JC, et al. 2010. The 1391
Surviving Sepsis Campaign: results of an international guideline-based performance 1392
improvement program targeting severe sepsis. Critical care medicine 38:367-74 1393
150. Liesenfeld O, Lehman L, Hunfeld KP, Kost G. 2014. Molecular diagnosis of sepsis: New aspects 1394
and recent developments. European journal of microbiology & immunology 4:1-25 1395
151. Lindholm L, Sarkkinen H. 2004. Direct identification of gram-positive cocci from routine blood 1396
cultures by using AccuProbe tests. Journal of clinical microbiology 42:5609-13 1397
152. Loonen AJ, de Jager CP, Tosserams J, Kusters R, Hilbink M, et al. 2014. Biomarkers and 1398
molecular analysis to improve bloodstream infection diagnostics in an emergency care unit. PloS 1399
one 9:e87315 1400
64
153. Lucignano B, Ranno S, Liesenfeld O, Pizzorno B, Putignani L, et al. 2011. Multiplex PCR allows 1401
rapid and accurate diagnosis of bloodstream infections in newborns and children with suspected 1402
sepsis. Journal of clinical microbiology 49:2252-8 1403
154. Mancini N, Clerici D, Diotti R, Perotti M, Ghidoli N, et al. 2008. Molecular diagnosis of sepsis in 1404
neutropenic patients with haematological malignancies. Journal of medical microbiology 57:601-1405
4 1406
155. Mancini N, Infurnari L, Ghidoli N, Valzano G, Clementi N, et al. 2014. Potential impact of a 1407
microarray-based nucleic acid assay for rapid detection of Gram-negative bacteria and resistance 1408
markers in positive blood cultures. Journal of clinical microbiology 52:1242-5 1409
156. March-Rossello GA, Munoz-Moreno MF, Garcia-Loygorri-Jordan de Urries MC, Bratos-Perez 1410
MA. 2013. A differential centrifugation protocol and validation criterion for enhancing mass 1411
spectrometry (MALDI-TOF) results in microbial identification using blood culture growth 1412
bottles. European journal of clinical microbiology & infectious diseases 32:699-704 1413
157. Marshall JC, Dellinger RP, Levy M. 2010. The Surviving Sepsis Campaign: a history and a 1414
perspective. Surgical infections 11:275-81 1415
158. Martinez RM, Bauerle ER, Fang FC, Butler-Wu SM. 2014. Evaluation of three rapid diagnostic 1416
methods for direct identification of microorganisms in positive blood cultures. Journal of clinical 1417
microbiology 52:2521-9 1418
159. Massire C, Gertz RE, Jr., Svoboda P, Levert K, Reed MS, et al. 2012. Concurrent serotyping and 1419
genotyping of pneumococci by use of PCR and electrospray ionization mass spectrometry. 1420
Journal of clinical microbiology 50:2018-25 1421
65
160. Massire C, Ivy CA, Lovari R, Kurepina N, Li H, et al. 2011. Simultaneous identification of 1422
mycobacterial isolates to the species level and determination of tuberculosis drug resistance by 1423
PCR followed by electrospray ionization mass spectrometry. Journal of clinical microbiology 1424
49:908-17 1425
161. Mauro MV, Cavalcanti P, Perugini D, Noto A, Sperli D, Giraldi C. 2012. Diagnostic utility of 1426
LightCycler SeptiFast and procalcitonin assays in the diagnosis of bloodstream infection in 1427
immunocompromised patients. Diagnostic microbiology and infectious disease 73:308-11 1428
162. McLaughlin RW, Vali H, Lau PC, Palfree RG, De Ciccio A, et al. 2002. Are there naturally 1429
occurring pleomorphic bacteria in the blood of healthy humans? Journal of clinical microbiology 1430
40:4771-5 1431
163. Meex C, Neuville F, Descy J, Huynen P, Hayette MP, et al. 2012. Direct identification of bacteria 1432
from BacT/ALERT anaerobic positive blood cultures by MALDI-TOF MS: MALDI Sepsityper 1433
kit versus an in-house saponin method for bacterial extraction. Journal of medical microbiology 1434
61:1511-6 1435
164. Mencacci A, Leli C, Cardaccia A, Montagna P, Moretti A, et al. 2011. Comparison of 1436
conventional culture with SeptiFast real-time PCR for microbial pathogen detection in clinical 1437
specimens other than blood. Journal of medical microbiology 60:1774-8 1438
165. Mestas J, Polanco CM, Felsenstein S, Dien Bard J. 2014. Performance of the Verigene Gram-1439
positive blood culture assay for direct detection of Gram-positive organisms and resistance 1440
markers in a pediatric hospital. Journal of clinical microbiology 52:283-7 1441
166. Metzgar D, Frinder M, Lovari R, Toleno D, Massire C, et al. 2013. Broad-spectrum biosensor 1442
capable of detecting and identifying diverse bacterial and Candida species in blood. Journal of 1443
clinical microbiology 51:2670-8 1444
66
167. Morgan MA, Marlowe E, Novak-Weekly S, Miller JM, Painter TM, et al. 2011. A 1.5 hour 1445
procedure for identification of Enterococcus Species directly from blood cultures. Journal of 1446
visualized experiments : JoVE 1447
168. Moussaoui W, Jaulhac B, Hoffmann AM, Ludes B, Kostrzewa M, et al. 2010. Matrix-assisted 1448
laser desorption ionization time-of-flight mass spectrometry identifies 90% of bacteria directly 1449
from blood culture vials. Clinical microbiology and infection 16:1631-8 1450
169. Naber CK. 2009. Staphylococcus aureus bacteremia: epidemiology, pathophysiology, and 1451
management strategies. Clinical infectious diseases 48 Suppl 4:S231-7 1452
170. Naber CK, Baddour LM, Giamarellos-Bourboulis EJ, Gould IM, Herrmann M, et al. 2009. 1453
Clinical consensus conference: survey on Gram-positive bloodstream infections with a focus on 1454
Staphylococcus aureus. Clinical infectious diseases 48 Suppl 4:S260-70 1455
171. Nagalingam S, Lisgaris M, Rodriguez B, Jacobs MR, Lederman M, et al. 2014. Identification of 1456
Occult Fusobacterium nucleatum Central Nervous System Infection Using PCR Electrospray 1457
Ionization Mass Spectrometry (PCR/ESI-MS). Journal of clinical microbiology 1458
172. Navarro E, Segura JC, Castano MJ, Solera J. 2006. Use of real-time quantitative polymerase 1459
chain reaction to monitor the evolution of Brucella melitensis DNA load during therapy and post-1460
therapy follow-up in patients with brucellosis. Clinical infectious diseases 42:1266-73 1461
173. Oliveira K, Procop GW, Wilson D, Coull J, Stender H. 2002. Rapid identification of 1462
Staphylococcus aureus directly from blood cultures by fluorescence in situ hybridization with 1463
peptide nucleic acid probes. Journal of clinical microbiology 40:247-51 1464
174. Opota O, Croxatto A, Prod'hom G, Greub G. 2015. Blood Culture-Based Diagnosis of 1465
Bacteremia: State of the Art. Clinical microbiology and infection 1466
67
175. Opota O, Jaton K, Greub G. 2015. Microbial diagnosis of bloodstream infection: towards 1467
molecular diagnosis directly from blood Clinical Microbiology and Infection 1468
176. Opota O, Ney B, Zanetti G, Jaton K, Greub G, Prod'hom G. 2014. Bacteremia caused by 1469
Comamonas kerstersii in a patient with diverticulosis. Journal of clinical microbiology 52:1009-1470
12 1471
177. Orszag P, Disque C, Keim S, Lorenz MG, Wiesner O, et al. 2014. Monitoring of patients 1472
supported by extracorporeal membrane oxygenation for systemic infections by broad-range rRNA 1473
gene PCR amplification and sequence analysis. Journal of clinical microbiology 52:307-11 1474
178. Osborn TM, Phillips G, Lemeshow S, Townsend S, Schorr CA, et al. 2014. Sepsis Severity 1475
Score: An Internationally Derived Scoring System From the Surviving Sepsis Campaign 1476
Database. Critical care medicine 1477
179. Palka-Santini M, Putzfeld S, Cleven BE, Kronke M, Krut O. 2007. Rapid identification, virulence 1478
analysis and resistance profiling of Staphylococcus aureus by gene segment-based DNA 1479
microarrays: application to blood culture post-processing. Journal of microbiological methods 1480
68:468-77 1481
180. Palomares JC, Bernal S, Marin M, Holgado VP, Castro C, et al. 2011. Molecular diagnosis of 1482
Aspergillus fumigatus endocarditis. Diagnostic microbiology and infectious disease 70:534-7 1483
181. Paolucci M, Foschi C, Tamburini MV, Ambretti S, Lazzarotto T, Landini MP. 2014. Comparison 1484
between MALDI-TOF MS and FilmArray Blood Culture Identification panel for rapid 1485
identification of yeast from positive blood culture. Journal of microbiological methods 104:92-3 1486
68
182. Paolucci M, Stanzani M, Melchionda F, Tolomelli G, Castellani G, et al. 2013. Routine use of a 1487
real-time polymerase chain reaction method for detection of bloodstream infections in 1488
neutropaenic patients. Diagnostic microbiology and infectious disease 75:130-4 1489
183. Parcell BJ, Orange GV. 2013. PNA-FISH assays for early targeted bacteraemia treatment. 1490
Journal of microbiological methods 95:253-5 1491
184. Pardo J, Klinker KP, Borgert SJ, Butler BM, Rand KH, Iovine NM. 2014. Detection of Neisseria 1492
meningitidis from negative blood cultures and cerebrospinal fluid with the FilmArray blood 1493
culture identification panel. Journal of clinical microbiology 52:2262-4 1494
185. Park Y, Kim BS, Choi KH, Shin DH, Lee MJ, et al. 2012. A novel multiplex real-time PCR assay 1495
for the concurrent detection of hepatitis A, B and C viruses in patients with acute hepatitis. PloS 1496
one 7:e49106 1497
186. Pasqualini L, Mencacci A, Leli C, Montagna P, Cardaccia A, et al. 2012. Diagnostic performance 1498
of a multiple real-time PCR assay in patients with suspected sepsis hospitalized in an internal 1499
medicine ward. Journal of clinical microbiology 50:1285-8 1500
187. Paul M, Kariv G, Goldberg E, Raskin M, Shaked H, et al. 2010. Importance of appropriate 1501
empirical antibiotic therapy for methicillin-resistant Staphylococcus aureus bacteraemia. The 1502
Journal of antimicrobial chemotherapy 65:2658-65 1503
188. Paul M, Shani V, Muchtar E, Kariv G, Robenshtok E, Leibovici L. 2010. Systematic review and 1504
meta-analysis of the efficacy of appropriate empiric antibiotic therapy for sepsis. Antimicrobial 1505
agents and chemotherapy 54:4851-63 1506
69
189. Peleg AY, Tilahun Y, Fiandaca MJ, D'Agata EM, Venkataraman L, et al. 2009. Utility of peptide 1507
nucleic acid fluorescence in situ hybridization for rapid detection of Acinetobacter spp. and 1508
Pseudomonas aeruginosa. Journal of clinical microbiology 47:830-2 1509
190. Pierce SE, Bell RL, Hellberg RS, Cheng CM, Chen KS, et al. 2012. Detection and Identification 1510
of Salmonella enterica, Escherichia coli, and Shigella spp. via PCR-electrospray ionization mass 1511
spectrometry: isolate testing and analysis of food samples. Applied and environmental 1512
microbiology 78:8403-11 1513
191. Pingle MR, Granger K, Feinberg P, Shatsky R, Sterling B, et al. 2007. Multiplexed identification 1514
of blood-borne bacterial pathogens by use of a novel 16S rRNA gene PCR-ligase detection 1515
reaction-capillary electrophoresis assay. Journal of clinical microbiology 45:1927-35 1516
192. Pkhakadze T, Dmitrienko OA. 2013. [The evaluation of effectiveness of application of automated 1517
system of polymerase chain reaction to detect meticillin-resistant and meticillin-sensitive 1518
Staphylococcus aureus in clinical material from trauma orthopedic patients]. Klinicheskaia 1519
laboratornaia diagnostika:48-51 1520
193. Prod'hom G, Bizzini A, Durussel C, Bille J, Greub G. 2010. Matrix-assisted laser desorption 1521
ionization-time of flight mass spectrometry for direct bacterial identification from positive blood 1522
culture pellets. Journal of clinical microbiology 48:1481-3 1523
194. Prod'hom G, Durussel C, Greub G. 2013. A simple blood-culture bacterial pellet preparation for 1524
faster accurate direct bacterial identification and antibiotic susceptibility testing with the VITEK 1525
2 system. Journal of medical microbiology 62:773-7 1526
195. Radstrom P, Knutsson R, Wolffs P, Lovenklev M, Lofstrom C. 2004. Pre-PCR processing: 1527
strategies to generate PCR-compatible samples. Molecular biotechnology 26:133-46 1528
70
196. Rand KH, Delano JP. 2014. Direct identification of bacteria in positive blood cultures: 1529
comparison of two rapid methods, FilmArray and mass spectrometry. Diagnostic microbiology 1530
and infectious disease 79:293-7 1531
197. Rath PM, Saner F, Paul A, Lehmann N, Steinmann E, et al. 2012. Multiplex PCR for rapid and 1532
improved diagnosis of bloodstream infections in liver transplant recipients. Journal of clinical 1533
microbiology 50:2069-71 1534
198. Rello J, Lisboa T, Lujan M, Gallego M, Kee C, et al. 2009. Severity of pneumococcal pneumonia 1535
associated with genomic bacterial load. Chest 136:832-40 1536
199. Rigby S, Procop GW, Haase G, Wilson D, Hall G, et al. 2002. Fluorescence in situ hybridization 1537
with peptide nucleic acid probes for rapid identification of Candida albicans directly from blood 1538
culture bottles. Journal of clinical microbiology 40:2182-6 1539
200. Rogina P, Skvarc M, Stubljar D, Kofol R, Kaasch A. 2014. Diagnostic Utility of Broad Range 1540
Bacterial 16S rRNA Gene PCR with Degradation of Human and Free Bacterial DNA in 1541
Bloodstream Infection Is More Sensitive Than an In-House Developed PCR without Degradation 1542
of Human and Free Bacterial DNA. Mediators of inflammation 2014:108592 1543
201. Ryzhov V, Fenselau C. 2001. Characterization of the protein subset desorbed by MALDI from 1544
whole bacterial cells. Analytical chemistry 73:746-50 1545
202. Sachse S, Straube E, Lehmann M, Bauer M, Russwurm S, Schmidt KH. 2009. Truncated human 1546
cytidylate-phosphate-deoxyguanylate-binding protein for improved nucleic acid amplification 1547
technique-based detection of bacterial species in human samples. Journal of clinical microbiology 1548
47:1050-7 1549
71
203. Sampath R, Hall TA, Massire C, Li F, Blyn LB, et al. 2007. Rapid identification of emerging 1550
infectious agents using PCR and electrospray ionization mass spectrometry. Annals of the New 1551
York Academy of Sciences 1102:109-20 1552
204. Sampath R, Hofstadler SA, Blyn LB, Eshoo MW, Hall TA, et al. 2005. Rapid identification of 1553
emerging pathogens: coronavirus. Emerging infectious diseases 11:373-9 1554
205. Sampath R, Mulholland N, Blyn LB, Massire C, Whitehouse CA, et al. 2012. Comprehensive 1555
biothreat cluster identification by PCR/electrospray-ionization mass spectrometry. PloS one 1556
7:e36528 1557
206. Samuel LP, Tibbetts RJ, Agotesku A, Fey M, Hensley R, Meier FA. 2013. Evaluation of a 1558
microarray-based assay for rapid identification of Gram-positive organisms and resistance 1559
markers in positive blood cultures. Journal of clinical microbiology 51:1188-92 1560
207. Sarovich DS, Colman RE, Price EP, Massire C, Von Schulze AT, et al. 2013. Molecular 1561
genotyping of Acinetobacter spp. isolated in Arizona, USA, using multilocus PCR and mass 1562
spectrometry. Journal of medical microbiology 62:1295-300 1563
208. Scanvic A, Courdavault L, Sollet JP, Le Turdu F. 2011. [Interest of real-time PCR Xpert 1564
MRSA/SA on GeneXpert((R)) DX System in the investigation of staphylococcal bacteremia]. 1565
Pathologie-biologie 59:67-72 1566
209. Schaub N, Boldanova T, Noveanu M, Arenja N, Hermann H, et al. 2014. Incremental value of 1567
multiplex real-time PCR for the early diagnosis of sepsis in the emergency department. Swiss 1568
medical weekly 144:w13911 1569
72
210. Schaumann R, Knoop N, Genzel GH, Losensky K, Rosenkranz C, et al. 2013. Discrimination of 1570
Enterobacteriaceae and Non-fermenting Gram Negative Bacilli by MALDI-TOF Mass 1571
Spectrometry. The open microbiology journal 7:118-22 1572
211. Schmittgen TD, Livak KJ. 2008. Analyzing real-time PCR data by the comparative C(T) method. 1573
Nature protocols 3:1101-8 1574
212. Schreiber J, Nierhaus A, Braune SA, de Heer G, Kluge S. 2013. Comparison of three different 1575
commercial PCR assays for the detection of pathogens in critically ill sepsis patients. 1576
Medizinische Klinik, Intensivmedizin und Notfallmedizin 108:311-8 1577
213. Seng P, Abat C, Rolain JM, Colson P, Lagier JC, et al. 2013. Identification of rare pathogenic 1578
bacteria in a clinical microbiology laboratory: impact of matrix-assisted laser desorption 1579
ionization-time of flight mass spectrometry. Journal of clinical microbiology 51:2182-94 1580
214. Shen J, Wang F, Li F, Housley R, Carolan H, et al. 2013. Rapid identification and differentiation 1581
of non-O157 Shiga toxin-producing Escherichia coli using polymerase chain reaction coupled to 1582
electrospray ionization mass spectrometry. Foodborne pathogens and disease 10:737-43 1583
215. Shepard JR, Addison RM, Alexander BD, Della-Latta P, Gherna M, et al. 2008. Multicenter 1584
evaluation of the Candida albicans/Candida glabrata peptide nucleic acid fluorescent in situ 1585
hybridization method for simultaneous dual-color identification of C. albicans and C. glabrata 1586
directly from blood culture bottles. Journal of clinical microbiology 46:50-5 1587
216. Simner PJ, Buckwalter SP, Uhl JR, Wengenack NL, Pritt BS. 2013. Detection and identification 1588
of yeasts from formalin-fixed, paraffin-embedded tissue by use of PCR-electrospray ionization 1589
mass spectrometry. Journal of clinical microbiology 51:3731-4 1590
73
217. Simner PJ, Uhl JR, Hall L, Weber MM, Walchak RC, et al. 2013. Broad-range direct detection 1591
and identification of fungi by use of the PLEX-ID PCR-electrospray ionization mass 1592
spectrometry (ESI-MS) system. Journal of clinical microbiology 51:1699-706 1593
218. Sitnik R, Marra AR, Petroni RC, Ramos OP, Martino MD, et al. 2014. SeptiFast for diagnosis of 1594
sepsis in severely ill patients from a Brazilian hospital. Einstein 12:191-6 1595
219. Snyder JW, Munier GK, Heckman SA, Camp P, Overman TL. 2009. Failure of the BD GeneOhm 1596
StaphSR assay for direct detection of methicillin-resistant and methicillin-susceptible 1597
Staphylococcus aureus isolates in positive blood cultures collected in the United States. Journal 1598
of clinical microbiology 47:3747-8 1599
220. Sogaard M, Hansen DS, Fiandaca MJ, Stender H, Schonheyder HC. 2007. Peptide nucleic acid 1600
fluorescence in situ hybridization for rapid detection of Klebsiella pneumoniae from positive 1601
blood cultures. Journal of medical microbiology 56:914-7 1602
221. Spencer DH, Sellenriek P, Burnham CA. 2011. Validation and implementation of the GeneXpert 1603
MRSA/SA blood culture assay in a pediatric setting. American journal of clinical pathology 1604
136:690-4 1605
222. Stamper PD, Cai M, Howard T, Speser S, Carroll KC. 2007. Clinical validation of the molecular 1606
BD GeneOhm StaphSR assay for direct detection of Staphylococcus aureus and methicillin-1607
resistant Staphylococcus aureus in positive blood cultures. Journal of clinical microbiology 1608
45:2191-6 1609
223. Steensels D, Vankeerberghen A, De Beenhouwer H. 2013. Towards multitarget testing in 1610
molecular microbiology. International journal of microbiology 2013:121057 1611
74
224. Stevenson LG, Drake SK, Murray PR. 2010. Rapid identification of bacteria in positive blood 1612
culture broths by matrix-assisted laser desorption ionization-time of flight mass spectrometry. 1613
Journal of clinical microbiology 48:444-7 1614
225. Stone NR, Gorton RL, Barker K, Ramnarain P, Kibbler CC. 2013. Evaluation of PNA-FISH 1615
yeast traffic light for rapid identification of yeast directly from positive blood cultures and 1616
assessment of clinical impact. Journal of clinical microbiology 51:1301-2 1617
226. Sullivan KV, Deburger B, Roundtree SS, Ventrola CA, Blecker-Shelly DL, Mortensen JE. 2014. 1618
Pediatric multicenter evaluation of the verigene gram-negative blood culture test for rapid 1619
detection of inpatient bacteremia involving gram-negative organisms, extended-spectrum Beta-1620
lactamases, and carbapenemases. Journal of clinical microbiology 52:2416-21 1621
227. Sullivan KV, Turner NN, Roundtree SS, Young S, Brock-Haag CA, et al. 2013. Rapid detection 1622
of Gram-positive organisms by use of the Verigene Gram-positive blood culture nucleic acid test 1623
and the BacT/Alert Pediatric FAN system in a multicenter pediatric evaluation. Journal of 1624
clinical microbiology 51:3579-84 1625
228. Taliaferro LP, Galvin TA, Ma H, Shaheduzzaman S, Williams DK, et al. 2014. Evaluation of the 1626
broad-range PCR-electrospray ionization mass spectrometry (PCR/ESI-MS) system and virus 1627
microarrays for virus detection. Viruses 6:1876-96 1628
229. Tan TY, Ng LS, Kwang LL. 2008. Evaluation of disc susceptibility tests performed directly from 1629
positive blood cultures. Journal of clinical pathology 61:343-6 1630
230. Tang YW, Kilic A, Yang Q, McAllister SK, Li H, et al. 2007. StaphPlex system for rapid and 1631
simultaneous identification of antibiotic resistance determinants and Panton-Valentine leukocidin 1632
detection of staphylococci from positive blood cultures. Journal of clinical microbiology 1633
45:1867-73 1634
75
231. Tang YW, Lowery KS, Valsamakis A, Schaefer VC, Chappell JD, et al. 2013. Clinical accuracy 1635
of a PLEX-ID flu device for simultaneous detection and identification of influenza viruses A and 1636
B. Journal of clinical microbiology 51:40-5 1637
232. Taur Y, Cohen N, Dubnow S, Paskovaty A, Seo SK. 2010. Effect of antifungal therapy timing on 1638
mortality in cancer patients with candidemia. Antimicrobial agents and chemotherapy 54:184-90 1639
233. Tissari P, Zumla A, Tarkka E, Mero S, Savolainen L, et al. 2010. Accurate and rapid 1640
identification of bacterial species from positive blood cultures with a DNA-based microarray 1641
platform: an observational study. Lancet 375:224-30 1642
234. Tojo M, Fujita T, Ainoda Y, Nagamatsu M, Hayakawa K, et al. 2014. Evaluation of an automated 1643
rapid diagnostic assay for detection of Gram-negative bacteria and their drug-resistance genes in 1644
positive blood cultures. PloS one 9:e94064 1645
235. Valentine N, Wunschel S, Wunschel D, Petersen C, Wahl K. 2005. Effect of culture conditions 1646
on microorganism identification by matrix-assisted laser desorption ionization mass spectrometry. 1647
Applied and environmental microbiology 71:58-64 1648
236. Venkatesh M, Flores A, Luna RA, Versalovic J. 2010. Molecular microbiological methods in the 1649
diagnosis of neonatal sepsis. Expert review of anti-infective therapy 8:1037-48 1650
237. Vernon SD, Shukla SK, Conradt J, Unger ER, Reeves WC. 2002. Analysis of 16S rRNA gene 1651
sequences and circulating cell-free DNA from plasma of chronic fatigue syndrome and non-1652
fatigued subjects. BMC microbiology 2:39 1653
238. Vernon SD, Unger ER, Dimulescu IM, Rajeevan M, Reeves WC. 2002. Utility of the blood for 1654
gene expression profiling and biomarker discovery in chronic fatigue syndrome. Disease markers 1655
18:193-9 1656
76
239. Verroken A, Defourny L, Lechgar L, Magnette A, Delmee M, Glupczynski Y. 2014. Reducing 1657
time to identification of positive blood cultures with MALDI-TOF MS analysis after a 5-h 1658
subculture. European journal of clinical microbiology & infectious diseases 1659
240. Vogne C, Prod'Hom G, Jaton K, Decosterd L, Greub G. 2014. A simple, robust and rapid 1660
approach to detect carbapenemases in Gram negative isolates by MALDI-TOF mass 1661
spectrometry: validation with triple quadripole tandem mass spectrometry, microarray and PCR. 1662
Clinical microbiology and infection 1663
241. von Lilienfeld-Toal M, Lehmann LE, Raadts AD, Hahn-Ast C, Orlopp KS, et al. 2009. Utility of 1664
a commercially available multiplex real-time PCR assay to detect bacterial and fungal pathogens 1665
in febrile neutropenia. Journal of clinical microbiology 47:2405-10 1666
242. Wagner J, Schilcher G, Zollner-Schwetz I, Hoenigl M, Valentin T, et al. 2013. Microbiological 1667
screening for earlier detection of central venous catheter-related bloodstream infections. 1668
European journal of clinical investigation 43:964-9 1669
243. Wain J, Diep TS, Ho VA, Walsh AM, Nguyen TT, et al. 1998. Quantitation of bacteria in blood 1670
of typhoid fever patients and relationship between counts and clinical features, transmissibility, 1671
and antibiotic resistance. Journal of clinical microbiology 36:1683-7 1672
244. Wallet F, Nseir S, Baumann L, Herwegh S, Sendid B, et al. 2010. Preliminary clinical study using 1673
a multiplex real-time PCR test for the detection of bacterial and fungal DNA directly in blood. 1674
Clinical microbiology and infection 16:774-9 1675
245. Wang F, Massire C, Li H, Cummins LL, Li F, et al. 2011. Molecular characterization of drug-1676
resistant Mycobacterium tuberculosis isolates circulating in China by multilocus PCR and 1677
electrospray ionization mass spectrometry. Journal of clinical microbiology 49:2719-21 1678
77
246. Wang Z, Russon L, Li L, Roser DC, Long SR. 1998. Investigation of spectral reproducibility in 1679
direct analysis of bacteria proteins by matrix-assisted laser desorption/ionization time-of-flight 1680
mass spectrometry. Rapid communications in mass spectrometry : RCM 12:456-64 1681
247. Washington MA, Agee WA, 3rd, Kajiura L, Staege CM, Uyehara CF, Barnhill JC. 2014. An 1682
analysis of Staphylococcus aureus infections at a military medical center using the PLEX-ID 1683
combined polymerase chain reaction-mass spectrometry system. Military medicine 179:445-50 1684
248. Weigel KM, Jones KL, Do JS, Melton Witt J, Chung JH, et al. 2013. Molecular viability testing 1685
of bacterial pathogens from a complex human sample matrix. PloS one 8:e54886 1686
249. Weller SA, Cox V, Essex-Lopresti A, Hartley MG, Parsons TM, et al. 2012. Evaluation of two 1687
multiplex real-time PCR screening capabilities for the detection of Bacillus anthracis, Francisella 1688
tularensis and Yersinia pestis in blood samples generated from murine infection models. Journal 1689
of medical microbiology 61:1546-55 1690
250. Wellinghausen N, Kochem AJ, Disque C, Muhl H, Gebert S, et al. 2009. Diagnosis of bacteremia 1691
in whole-blood samples by use of a commercial universal 16S rRNA gene-based PCR and 1692
sequence analysis. Journal of clinical microbiology 47:2759-65 1693
251. Werner AS, Cobbs CG, Kaye D, Hook EW. 1967. Studies on the bacteremia of bacterial 1694
endocarditis. JAMA : the journal of the American Medical Association 202:199-203 1695
252. Westh H, Lisby G, Breysse F, Boddinghaus B, Chomarat M, et al. 2009. Multiplex real-time PCR 1696
and blood culture for identification of bloodstream pathogens in patients with suspected sepsis. 1697
Clinical microbiology and infection 15:544-51 1698
78
253. Wiesinger-Mayr H, Jordana-Lluch E, Martro E, Schoenthaler S, Noehammer C. 2011. 1699
Establishment of a semi-automated pathogen DNA isolation from whole blood and comparison 1700
with commercially available kits. Journal of microbiological methods 85:206-13 1701
254. Wiesinger-Mayr H, Vierlinger K, Pichler R, Kriegner A, Hirschl AM, et al. 2007. Identification 1702
of human pathogens isolated from blood using microarray hybridisation and signal pattern 1703
recognition. BMC microbiology 7:78 1704
255. Wilson DA, Joyce MJ, Hall LS, Reller LB, Roberts GD, et al. 2005. Multicenter evaluation of a 1705
Candida albicans peptide nucleic acid fluorescent in situ hybridization probe for characterization 1706
of yeast isolates from blood cultures. Journal of clinical microbiology 43:2909-12 1707
256. Wisplinghoff H, Bischoff T, Tallent SM, Seifert H, Wenzel RP, Edmond MB. 2004. Nosocomial 1708
bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide 1709
surveillance study. Clinical infectious diseases 39:309-17 1710
257. Wittwer CT, Herrmann MG, Gundry CN, Elenitoba-Johnson KS. 2001. Real-time multiplex PCR 1711
assays. Methods 25:430-42 1712
258. Wojewoda CM, Sercia L, Navas M, Tuohy M, Wilson D, et al. 2013. Evaluation of the Verigene 1713
Gram-positive blood culture nucleic acid test for rapid detection of bacteria and resistance 1714
determinants. Journal of clinical microbiology 51:2072-6 1715
259. Wolffs P, Grage H, Hagberg O, Radstrom P. 2004. Impact of DNA polymerases and their buffer 1716
systems on quantitative real-time PCR. Journal of clinical microbiology 42:408-11 1717
260. Wolk D, Mitchell S, Patel R. 2001. Principles of molecular microbiology testing methods. 1718
Infectious disease clinics of North America 15:1157-204 1719
79
261. Wolk DM, Kaleta EJ, Wysocki VH. 2012. PCR-electrospray ionization mass spectrometry: the 1720
potential to change infectious disease diagnostics in clinical and public health laboratories. The 1721
Journal of molecular diagnostics : JMD 14:295-304 1722
262. Wortmann G, Weintrob A, Barber M, Scott P, Zoll ST, et al. 2008. Genotypic evolution of 1723
Acinetobacter baumannii strains in an outbreak associated with war trauma. Infection control and 1724
hospital epidemiology : the official journal of the Society of Hospital Epidemiologists of America 1725
29:553-5 1726
263. Wyllie DH, Crook DW, Peto TE. 2006. Mortality after Staphylococcus aureus bacteraemia in two 1727
hospitals in Oxfordshire, 1997-2003: cohort study. Bmj 333:281 1728
264. Yoo SM, Choi JY, Yun JK, Choi JK, Shin SY, et al. 2010. DNA microarray-based identification 1729
of bacterial and fungal pathogens in bloodstream infections. Molecular and cellular probes 1730
24:44-52 1731
265. Zaragoza R, Artero A, Camarena JJ, Sancho S, Gonzalez R, Nogueira JM. 2003. The influence of 1732
inadequate empirical antimicrobial treatment on patients with bloodstream infections in an 1733
intensive care unit. Clinical microbiology and infection 9:412-8 1734
266. Ziegler I, Josefson P, Olcen P, Molling P, Stralin K. 2014. Quantitative data from the SeptiFast 1735
real-time PCR is associated with disease severity in patients with sepsis. BMC infectious diseases 1736
14:155 1737
1738
1739
80
TABLES AND FIGURES 1740
1741
1742
81
Table 1: Commercially available molecular systems for the microbial identification of pathogen 1743
during BSI. Adapted from (174; 175) 1744
1745
SYSTEM (MANUFACTURER)
METHODS TIME TO RESULT
MICROORGANISM COVERAGE
RESISTANCE AND VIRULENCE MARKERS
SENSITIVITY, SPECIFICITY AND CORRELATION WITH CONVENTIONAL METHODS
REFERENCES
IDENTIFICATION FROM POSITIVE BLOOD-CULTURE
PNA FISH and QuickFISH (AdvanDx, Wolburn, MA)
FISH <1-3 hours 4 gram-negative 4 gram-negative 5 fungi
none 97-100% 90-100% 96-99%
(31; 55; 56; 85; 103; 106; 167; 220; 225)
AccuProbe
(Gen-Probe, San Diego, CA, USA)
FISH <1 hour S. aureus Enterococcus spp. S. pneumoniae Streptococcus group A Streptococcus group B
none 80.8-100% 98.7-100%
nr
(9; 151)
Verigene (Nanosphere, Northbrook, IL, USA)
Microarray 2.5 hours 12 gram-negative 9 gram-negative
mecA, vanA/B, blaKPC , blaNDM, CTX-M,VIM,IMP,OXA12
81-100% 98-100%
nr
(5; 14; 17; 26; 60; 155; 165; 226; 227; 234)
Prove-it Sepsis (Mobidiag, Finland)
Microarray 3.5 hours 60 bacteria 13 fungi
mecA 95 % 99 %
nr
(233)
FilmArray (Idaho Technology, Salt Lake City, UT, USA)
Multiplex PCR 1 hour 8 gram-negative 11 gram-negative 5 fungi
mecA, vanA/B, blaKPC
97-95% 91-98%
nr
(6; 19; 181; 196)
Xpert MRSA/SA BC (Cepheid, Sunnyvale, CA, USA)
Real-time PCR 1 hour S. aureus mecA 100% 99-100%
nr
(45; 54; 140; 208; 221)
StaphSR assay (BD GeneOhm, San Diego, CA, USA)
Multiplex PCR 1-2 hours S. aureus mecA 96-100% 95-98%
nr
(97; 130)
StaphPlex (Genaco Biomedical Products, Huntsville, AL, USA)
Multiplex PCR + Microarray
5 hours S. aureus mecA (+ PVL) 100% 95-100%
92%
(230)
MALDI-TOF MS Bruker Daltonics (Bremen, Germany) or bioMérieux (Marcy l’Etoile, France)
Mass-spectrometry
<1 hour <1000a not in routine - -
76-99%
(42; 52; 156; 158; 193; 194; 224)
IDENTIFICATION FROM WHOLE BLOOD
SepsiTest (Molzym, Bremen, Germany)
Broad range PCR + sequencing
6 hours (1-10b ml)
>345 bacteria and fungi none 21-87% 85-96%
nr
(145; 152; 177; 200; 212; 250)
SeptiFast (Roche Molecular System, Germany )
Multiple broad-range real time PCR
3.5-5 hours (1.5 ml)
6 Gram positive 8 gram-negative 5 fungi
mecA b 43-95% 60-100% 43-83%
(25; 35; 53; 80; 95; 124; 127; 139; 143; 145-147; 153; 154; 161; 164; 182; 186; 197; 244)
MagicPlex (Seegene, Seoul, Korea)
Multiple PCR+ multiplex real time PCR
3-5 hours (1ml)
21 bacteria at species level (90 at genus level) 6 fungi
mecA, vanA/B 37-65% 77-92%
73%
(30; 152)
VYOO (SIRS-Lab, Jena, Germany)
Multiplex PCR + electrophoresis
8 hours (5ml)
14 Gram positive 18 gram-negative 7 fungi
none nr nr
70%
(20; 84; 212)
PLEX-ID
(Abbott Molecular, Des Plaines, IL)Laboratories, IL, USA)
Multiplex broad-range PCR + ESI-MS
6 hours (1.25-5 c ml)
up to 800 (Gram positive, Gram negative, fungi)
mecA, blaKPC , vanA/B
50-91%d 98-99% 79-97%
(13; 166)
82
a Dependant on the mass-spectrum database 1746 b With an additional kit 1747 c For the latest version 1748 d depending on the volume 1749 nr=non reported 1750
1751
83
1752
Figure 1: Different molecular technologies used for the detection and identification of microbes 1753
during bloodstream infections. Molecular methods for the diagnosis of BSI include nucleic acid-based 1754
methods and non-nucleic acid-based. Nucleic acid amplification-based techniques can be applied on 1755
positive blood cultures or used directly on blood whereas non nucleic acid amplification-based techniques 1756
such as FISH (fluorescent in situ hybridization) and microarray or non-nucleic acid based methods such 1757
as MALDI-TOF MS (Matrix-Assisted Laser Desorption Ionization Time-Of-Flight Mass Spectrometry) 1758
can be used only on positive blood culture. Adapted from (150). 1759
1760
84
1761
1762
Figure 2: Conventional and molecular methods for the identification of microorganisms from 1763
positive blood culture. When a blood culture is detected as positive several strategies are available to 1764
identify the incriminated microorganism after the initial Gram staining: (1) Identification directly from 1765
the positive blood culture using nucleic acid-based methods, (2) Identification after microbe’s enrichment, 1766
namely a purified bacterial pellet suitable for MALDI-TOF MS analysis and some automated 1767
identification/AST approaches such as Vitek2 and Phoenix and rapid biochemical tests HMRZ and 1768
ESBLNP (3) Identification after a subculture; to date the automated system Vitek2 has been validated on 1769
short subculture as well as the rapid biochemical tests HMRZ and CARNP; long subculture are suitable 1770
for any type of analysis including phenotypic characterization, automated systems such as Vitek2, 1771
Phoenix and MicroScan WalkAway as well as PCR followed by sequencing. Adapted from (174). 1772
85
1773
1774
Figure 3: Microbial identification from positive blood culture using MALDI-TOF MS. The sample 1775
that is deposited on the MALDI-TOF multi-well plate can be a bacterial pellet obtained by centrifugation 1776
and erythrocytes lysis. Alternatively, the sample can be bacteria obtained after a subculture. Proteins are 1777
then extracted and embedded in a matrix directly on the multi-well conductive metal plate and submitted 1778
to the MALDI-TOF MS, which separate the proteins according to their MW and their charges. This 1779
generates a mass spectrum (MS) which, in a second stage is compared to a database of spectra. This 1780
analysis provides the identification of the microorganism with a confidence score, which allows 1781
acceptation at the species or at the genus level. The identification is interpreted according to the Gram 1782
staining and to the clinical presentation. Adapted from (52). 1783
86
1784
1785
1786
Figure 4: Schematic workflow of PCR/ESI-MS. (a) Nucleic-acids are extracted from the sample, 1787
directly from whole blood, and (b) amplified by multiple PCR using multiple pairs of primers; each 1788
colour represents a different primer. After amplification, the molecular mass of the amplicon(s) is 1789
precisely determined using ESI-MS (c, d and e), from which (f) the base composition of the amplicon(s) 1790
is deduced. Finally, (g) informations - base composition - obtained from one or more amplicon(s) are 1791
compared with a database, which provides the identification with a confidence score. Adapted from (123). 1792
1793